Mania Africa 💎🌍

Nvidia and OpenAI abandon unfinished $100B deal in favour of $30B investment
View Article | Join the HN Conversation

Summary of HN discussion 🧵👇

ハントレスとマイクロソフトがClickFix攻撃の継続的な人気と進化について詳細に説明

Huntress, Microsoft Detail the Continued Popularity, Evolution of ClickFix Attacks #SecurityBoulevard (Feb 19)

securityboulevard.com/2026/02/hunt...

thehill.com/policy/techn...

DOL Announces New Community College Grant Aimed at Creating Workforce Programs and Systems

Details for how community colleges can apply are available online, with an application deadline of May 20: https://ow.ly/c6fW50YiZs7

Moscow [Augusta Co, VA] Law Enforcement reports Non-Tstm Wnd Dmg at 20 Feb, 5:10 PM EST -- Four reports of trees down in Augusta County around 5 PM.
at Sat, 21 Feb 2026 00:03:30 +0000 via IEMbot
Additional Details Here.

@wired.com
They asked technologists, activists, aid groups, and cybersecurity experts for guidance on how to organize and collaborate in an age of surveillance. Here’s the result: www.wired.com/story/how-to...

New AI Data Leaks—More Than 1 Billion IDs And Photos Exposed

buff.ly/tZPfqE0

#news #tech #technology #AI #deepseek #openai #chatgpt #grok #aislop #nvidia #leak #security #privacy

Every Claude Code and ChatGPT subscription has Anthropic or OpenAI spending anywhere from $5 to $13 to make $1 when it comes to these coding products. To not include this is journalistic malpractice and to act as if this will go away is too. Disgraceful.
bsky.app/profile/edzi...

Russia stepping up hybrid attacks, preparing for long standoff with West, Dutch intelligence warns therecord.media/russia-cyber...

The Tumbler Ridge shooter had used OpenAI to discuss scenarios involving gun violence.

It was bad enough to get flagged and reviewed by employees, who warned the company & suggested alerting law enforcement.

The company said No.

This tragedy could have been prevented.

Gut wrenching.

Sam Altman claims ChatGPT's adult mode will 'be able to safely relax the restrictions ...
->TechRadar | More on "ChatGPT adult mode and safety" at BigEarthData.ai | #Chatgpt #AI

“DHS Opens a Billion-Dollar Tab With [Peter Thiel’s] Palantir” 2/19/26 www.wired.com/story/depart...

"Ukrainian Hackers Reveal Russia Uses Belarusian Infrastructure to Guide Drones Against Ukraine and NATO"
united24media.com/latest-news/...

Runlayer is now offering secure OpenClaw agentic capabilities for large enterprises

OpenClaw, the open source AI agent that excels in autonomous tasks on computers and with which users can communicate through popular messaging apps, has undoubtedly become a phenomenon since its launch in November…

Ex #Google warns of Big Tech holding 1984-level control
#NEAR Protocol says #AI giants will manipulate users unless we build private, user-owned alternatives.
www.thestreet.com/crypto/marke... #OpenAI #Anthropic #Nearcon #Nvidia #Sony #FilterLabs #ShieldAI #FigureAI #BostonDynamics #FulcraDynamics

Colour me not sad.

The deserve it.

They’ve been robing us of our data for years.

thehill.com/policy/techn...

Sue this company out of existence.

www.wsj.com/us-news/law/...

Some good news
You can just do things in 2026 including making new models and running them on your PC and sharing them for free and not use provider for trivial tasks...

You know, the opposite of OpenAI and openclaw

Another reason to abandon google (use duckduckgo)for searches & email (use proton.me)

boingboing.net/2026/02/06/i...

As someone in tech who isn't AI-pilled, I hope the price hike / bubble pop comes sooner than later.

Right now it's like the early days of Uber. All the AI companies want that early market domination, but the best way to do that is bleed cash until they're one of the only players in town.

cybernews.com/privacy/de-g...

Keep Android Open

Please read this and help take action for the Keep Android Open movement. Tyranny has no place in the world, and corporate tyranny is no different ​:neofox_sign_no:​

.@PaloAltoNtwks acquires Koi, adds AI agent endpoint security bit.ly/3Omn6Tl Palo Alto Networks said it will acquire Koi in a move that will add endpoint security to AI agents.

Jolene Parish and co-founders Barret Zoph and Luke Metz have rejoined OpenAI from Mira Murati’s Thinking Machines Lab, a startup valued at $12 billion that recently launched its first AI product.

"Late last month, The Information reported that Apple was developing an AI wearable — an AirTag-sized pendant with cameras that could be pinned to a user’s shirt."


💄#SurveillanceChic 🤷🏻‍♀️


Are we cooked? 👩🏻‍🍳
techcrunch.com/2026/02/17/a...

Is Australia’s $100 million investment in Palantir beyond the pale? www.crikey.com.au/2026/02/20/c...
@albomp.bsky.social Madness. Strongly DISapprove.

open.substack.com/pub/leohohma...

  The latest episode of the IEA’s Everything Energy podcast surveys the global landscape for energy jobs, looking at the biggest engines for growth across the sector and the challenges that could undermine future momentum. Now available on Apple Podcasts and Spotify, the episode features insights…

Move beyond data monoliths with Da2a, an open-source agentic data platform using the A2A protocol for decentralized analytics. #softwarearchitecture

These Nazis are ruthless in their mission and are now looking at identifying people with untested and very flawed software. They will be shipping innocent people out of the country sooner rather than later! Im almost positive that they will also use it during the midterms.

Googleは2025年に175万件以上のPlayストアアプリの申請をブロックした

Google blocked over 1.75 million Play Store app submissions in 2025 #BleepingComputer (Feb 19)

www.bleepingcomputer.com/news/securit...

Java is the most widely used programming language for enterprise apps, Android development, & university CS courses. But if you have a Mac, Java doesn't come pre-installed. In this step-by-step guide Daniel shows you how to install Java on Mac to get you started.
www.freecodecamp.org/news/how-to-...

This post appeared under this Techmeme headline:

Must-Have Investments for Security-Aware & Productive Teams

Explore must-have investments that boost team security, productivity, and collaboration with the right tools, training, and infrastructure.
#hackernews #news

Yet another good reason not to use AI.

Key Takeaways & Critical Action Items

Sarvam launched the Indus chat application featuring a 105-billion-parameter model.
The software operates on the web. It functions on Apple devices. It runs on Android hardware.
India generates 5.8% of global traffic for Anthropic.

TabRep: Training Tabular Diffusion Models with a Simple and Effective Continuous Representation

Jacob Si, Zijing Ou, Mike Qu, Zhengrui Xiang, Yingzhen Li

Action editor: Jan-Willem van de Meent

https://openreview.net/forum?id=yRbtFEh2OP

#tabrep #sparse #tabular

Well, this is depressing. More taxpayer dollar spent on tech to oppress us.

what a cluster
https://www.wsj.com/us-news/law/openai-employees-raised-alarms-about-canada-shooting-suspect-months-ago-b585df62

Going live in 10 min: RCT1 scenario playthrough continues with Karts & Coasters (OpenRCT2 on Linux).
Chill build, go-karts + coasters, clean park design.
youtube.com/live/qlAfnJ4...

#RollerCoasterTycoon #OpenRCT2 #LinuxGaming #VTuber #KartsAndCoasters #RCT1 #RetroGaming #CozyGaming

OpenAI, JioHotstar Partner to Launch ChatGPT-Powered Content Discovery in India

JioHotstar has joined forces with OpenAI to bring a ChatGPT-branded discovery experience to what is India’s most widely used streaming service, enabling its nearly half a billion monthly users …
#chatgpt #discovery #gpt

This post appeared under this Techmeme headline:

"/e/OS, a privacy-focused mobile operating system created by an open source advocate and former Linux Mandrake founder, is drawing steady interest from users looking to reduce their reliance on big tech ecosystems."
#degoogle
cybernews.com/privacy/de-g...

Cloud security is shifting fast ☁️ Nick Cavalancia says "identity has replaced the perimeter entirely" & warns that AI-driven attacks are accelerating the threat landscape🤖
virtualizationreview.com/articles/202...

Employees at OpenAI urged leaders to alert the police, but they opted not to.

I scored 17/20 on the UNSW AI face-detection demo facetest.psy.unsw.edu.au/aifa... /via @stilgherrian.com The Weekly Cybers newsletter

More about the online footprint of the #tumblerridge #shooter

#openai #chatgpt

www.wsj.com/us-news/law/...

1-15 February 2026 Cyber Attacks Timeline www.hackmageddon.com/2026/02/18/1...

9to5google.com/2026/02/20/n...

Google could reportedly invest $100M in AI cloud operator Fluidstack siliconangle.com/2026/02/20/g...

“Claw” is becoming to agent tech what “podcasts” were to iPods. Good news for Anthropic I suspect.

Critical BeyondTrust Remote Support flaw abused by ransomware gangs

📌 New Course from Google

Google AI Professional Certificate --- Coursera Certification with Discount

🧠 Key Topics Covered:

🔹 AI Tools like Gemini & NotebookLM

🔹 Productivity with AI Workflows

🔹 Capstone Project with AI Studio

🌐 Explore Course Details: imp.i384100.net/c/3288150/37...

Apple's Low-Cost Colorful MacBook: All the Rumors - MacRumors

www.macrumors.com/2026/02/20/l...

Gemini 3.1 Pro brings advanced logic to developers and consumers

Dina Doll and Resident Legal Diva Melba Pearson break down three alarming reports on government surveillance and digital privacy — from tech companies allegedly turning over user data to DHS over posts about ICE, to federal phone surveillance in Minneapolis, and an open letter urging Silicon Valley

Pretty neat tutorial for learning caching of network data with Swift: developer.apple.com/tutorials/ap...

💡 OpenAI launches $7.5M AI alignment fund
• Funding The Alignment Project for independent AI safety research
• Focus on strengthening AI safety and security globally
#OpenAI #ChatGPT
https://openai.com/index/advancing-independent-research-ai-alignment

🤔 www.wsj.com/us-news/law/...

This is how systems fail survivors and shield the rich and powerful from accountability.

Don’t come for me, if you don’t come with facts.

www.wired.com/story/dhs-wa...

ICYMI:

www.pcworld.com/article/3063...

Bengaluru-based Sarvam, which is building AI models for local languages, launches its Indus chat app in beta, powered by its Sarvam 105B model (Jagmeet Singh/TechCrunch)

Main Link | Techmeme Permalink

Age verification vendor Persona left frontend exposed, researchers say www.malwarebytes.com/blog/news/20...

Meta patents AI that could keep you posting from beyond the grave www.malwarebytes.com/blog/ai/2026...

Job scam uses fake Google Forms site to harvest Google logins www.malwarebytes.com/blog/scams/2...

Android malware taps Gemini to navigate infected devices www.theregister.com/2026/02/19/g...

LOL #AIGenerated

I told Gemini to represent itself dreaming.

Gemini's meta-prompt in alt-text

Les advierto…puse de prueba a Claude.ai pagando $20/mes a traducir capítulos de un libro mío a un lenguaje que no conozco y en el segundo capítulo después de de 10000 palabras me dejó a pie por cinco horas…eso OpenAI no lo hace por el mismo precio…

OpenAI Deploys 200 Employees to Build AI Smart Devices www.pymnts.com/artificial-i...

You don't need a high-tech heist to lose your data. A single misconfigured setting or an accidental deletion can do the job just as well.

Read more 🔽
www.kelownanow.com/watercooler/...

AI vs AI: Defense Without Humans in the Loop www.pymnts.com/artificial-i...

Trump administration recruits tech bosses to train ‘elite’ government coders—The scheme has partnered with tech giants including Apple, Coinbase, Meta, Microsoft, Nvidia, OpenAI, xAI and Palantir. www.ft.com/content/7691... @joemillerjr.ft.com @financialtimes.com

www.wired.com/story/depart...

OpenClaw's rapid rise as an open-source AI agent turned dangerous due to security flaws, emergent harmful behaviours, and the inability of current regulations to keep pace. It highlights urgent needs for oversight and architectural redesigns.

Claude Code's compaction discards data that's still on disk

https://github.com/anthropics/claude-code/issues/26771

futurism.com/artificial-i...

Modern Best Practices for Web Security Using AI and Automation

AI and intelligent automation are transforming cybersecurity operations. From SOAR-driven incident response and UEBA anomaly detection to predictive vulnerability scoring, organizations can reduce alert fatigue, shrin…
#hackernews #news

You know, I wondered about this but never checked. That's dope.

Ukrainian Hackers Reveal Russia Uses Belarusian Infrastructure to Guide Drones Against Ukraine and NATO
->UNITED24 Media | More on "Russia Belarus drone infrastructure hacking" at BigEarthData.ai | #Hacker #Drone #RussiaUkraineWar #StandWithUkraine #NATO #Technology #RussiaUkraineWar

Attackers cloned Gemini with 100k prompts. Cost: ~$100. Your moat isn't the model, it's never been the model. Autonomous engineering value is in the system that validates, tests, and ships—not the autocomplete wrapper.

Leading Japanese semiconductor supplier responding to ransomware attack #cybersecurity #hacking #news #infosec #security #technology #privacy

"The move represents OpenAI's most aggressive bet yet on the idea that the future of AI isn't about what models can say, but what they can do."
venturebeat.com/technology/o...

Discover Network Shows How Sharing Data Can Turn the Tide on Fraud www.pymnts.com/news/securit...

UK sounds alarm on rising cyber risks to businesses

The UK government launched a national campaign urging businesses to strengthen basic cyber defenses. The initiative follows new figures highlighting the scale of the threat. Serious cyber incidents cost businesses an average of …
#hackernews #news

Japanese tech giant Advantest hit by ransomware attack www.bleepingcomputer.com/news/securit...

PromptSpy is the first known Android malware to use generative AI at runtime www.bleepingcomputer.com/news/securit...

Imagine, at a global scale, how many more instances around intent, planning, & radicalisation progress or pathways don't reach the threshold where law enforcement is notified... | OpenAI Employees Raised Alarms About Canada Shooting Suspect Months Ago www.wsj.com/us-news/law/...

How An AI Gmail Summary Agent Saves Me 35 Minutes a Day

An AI-powered Gmail Summary Agent built with n8n can filter promotions, group emails by category, extract action items, and deliver one clean HTML summary daily. Instead of drowning in threads, you get a focused report…
#hackernews #ml #openai

Claude Code 2.1.50 changes
github.com/anthropics/c...

According to Ransomware.live, anubis ransomware group has added Micaforce Technology (🇨🇳) to its victims.

Look: $600B in planned AI compute by 2030 won't just buy chips — it'll reroute capital, spike electricity demand, and concentrate who controls the industry.
Artifical Intelligence | Scary | found on CNA

Use 2FA, keep software updated, and monitor system logs to prevent AI model breaches.

Nvidia Nears $30 Billion OpenAI Stake, Drops $100B Infrastructure Pledge — Nvidia is reported to be close to investing up to $30 billion in OpenAI, replacing an earlier $100 billion infrastructure commitment and reshaping the companies' strategic partnership. The shift would reallocate…

On February 17, 2026, hackers exploited a compromised npm publish token to release a malicious version ([email protected]) of the Cline CLI package, affecting developers for 8 hours. The attack involved injecting a postinstall script that installed the legitimate OpenCLAW package.

Age Verification totally secure and trustable. Yeah... sure... if you want to be spied on! 🤬

Hackers Expose Age-Verification Software Powering Surveillance Web
www.therage.co/persona-age-...

AI’s ‘connective tissue’ is woefully insecure, Cisco warns www.cybersecuritydive.com/news/ai-agen...

JFrog down 25% after Claude Code Security launch - worse than CrowdStrike's 8% hit. Entire raised $60M seed 10 days before Anthropic moved into their space. Model providers absorbing dev tool functions quarterly. #AI #DevTools https://www.implicator.ai/jfrog-lost-25-in-one-day-ai-native-startups-hav

New filing: "Ziff Davis v. OpenAI (Ziff Davis sues OpenAI over training and outputs)"
Doc #432: Response in Opposition to Motion

Download PDF | View Full Case

#CL70338311

New filing: "Ziff Davis v. OpenAI (Ziff Davis sues OpenAI over training and outputs)"
Doc #431: Notice (Other)

Download PDF | View Full Case

#CL70338311

New filing: "Silverman v. OpenAI (Book authors sue over ChatGPT training)"
Doc #262: Response in Opposition to Motion

Download PDF | View Full Case

#CL69950496

New filing: "Chabon v. OpenAI (Book authors sue over ChatGPT training)"
Doc #216: Response in Opposition to Motion

Download PDF | View Full Case

#CL69920598

Amazon pushes back on Financial Times report blaming AI coding tools for AWS outages #Technology #Cybersecurity Other

Updated Post: 3-2-1 Backups Explained
#Guides

How to understand AI decision logic

Understand AI's 'black box' logic. Discover why Explainable AI builds trust, ensures fairness, and aids debugging. Learn more at www.reork.com.

Explainable AI, AI Black Box, AI Transparency, AI Bias, Responsible AI

Ever hear the story of the man who used a digital cloud connected butt plug that couldn't be removed without app permission of his partner?

It got hacked, and they held his butt for ransom

That is the new world: all tech is now a cloud connected butt plug

Claude Code's compaction discards data that's still on disk (https://news.ycombinator.com/item?id=47096210)

Excessive token usage in Claude Code

https://github.com/anthropics/claude-code/issues/16856

Pulse Secure network hacked via backdoor embedded in its VPN software

“A study led by MIT researchers found that agentic AI developers seldom publish detailed information about how these tools were tested for safety.”

Making frontier cybersecurity capabilities available to defenders


https://news.ycombinator.com/item?id=47091469

SPF Alignment: Why is it Important for Improving DMARC

Originally published at SPF Alignment: Why is it Important for Improving DMARC by EasyDMARC.

Every day, inboxes receive millions of emails that ...
#hackernews #news

Microslop Copliot apparently had a vulnerability that can read protected emails…

😬

https://cybernews.com/security/microsoft-copilot-confidential-email-data-leak/

theconversation.com/openai-has-d...

“A plethora of unlicensed or unsecured AI apps on the Google Play store for Android, including those marketed for identity verification and editing, have exposed billions of records and personal data, cybersecurity experts have confirmed.” mashable.com/article/andr...

No bid contracting is some deep state $hit.

"Jain suggests that the five-year agreement could allow [Palantir] to expand its reach across DHS into agencies like the US Secret Service ..., [FEMA], [TSA), and the Cybersecurity and Infrastructure Security Agency (CISA)."

www.wired.com/story/depart...

www.wired.com/story/opencl...

I'm glad a pub with some reach is talking about this, finally

📰 Gemini 3.1 Pro Generates Complex Isometric SVG Scene, Showcasing AI’s Creative Leap

A Reddit user demonstrated that Google’s Gemini 3.1 Pro can generate a fully functional isometric 3D scene using only SVG code — no external tools or manual rendering. The feat, validat...

#AINews #AI #Teknoloji

Lessons From AI Hacking: Every Model, Every Layer Is Risky #cybersecurity #hacking #news #infosec #security #technology #privacy

If a parent can be held responsible, so can a multi billion dollar tech corporation

Wow. Guess which AI subscription I just cancelled.

It would have saved lives and had them be alert of the deadly situation.

Grateful to the Stanford Democrats & Stanford Speakers Bureau for hosting me and Bernie Sanders tonight at Stanford University for a townhall on the future of AI.

The Discord situation just keeps getting 'better'.

www.malwarebytes.com/blog/news/20...

Security researchers found Moltbook's AI social network has fundamental flaws, including security vulnerabilities and manipulated engagement numbers. #AI

Read more below.

Even the OpenAI CEO shares the skepticism over whether AI is truly responsible for widespread layoffs. f-st.co/3naeCkB

Nvidia is reportedly close to investing $30 billion in OpenAI as part of a larger funding round. #AI

Read more below.

Anthropic is profitable in a debt-driven-reverse-ponzi-scheme sort of way

Unit 42 2026 IR: Attackers exfil in 72 mins, ID flaws in 90% cases. AI speeds it 4x. Mfg peeps: Patch playbooks for minutes, not days. #CyberRisk #Manufacturing #OTSec https://jmfj.link/NUgMmP

"Is this vulnerable?" is the wrong question. "How fast can we upgrade?" is the right one.

Reachability is noisy. High-Velocity Hygiene is the future. Read more from our VP Security, Josh Bressers: https://anchore.com/blog/no-crystal-ball-but-2026-directions/

#DevSecOps #SoftwareSupplyChain

OpenAI’s president is a Trump mega-donor.

Greg Brockman said he started ‘getting involved politically’ in 2025.
www.theverge.com/ai-artificia...

New filing: "New York Times v. Microsoft (AI copyright lawsuit)"
Doc #1248: LETTER MOTION to Seal addressed to Magistrate Judge Ona T. Wang from Caitlin Sinclaire Blythe, Christopher S. Sun, and Elana Nightingale Dawson…

[full entry below 👇]

Download PDF | View Full Case

#CL68117049

techcrunch.com/2026/02/19/f...

I can honestly say, I've never thought to myself that I need the speaker on my kitchen to have a camera on it

Who's this product for?

www.engadget.com/ai/openai-wi...

"In its statement, OpenAI said after learning of the shooting, it reached out to RCMP with information on Van Rootselaar and her use of ChatGPT.

An RCMP spokesperson confirmed to CBC News that the platform reached out after the shooting, but said OpenAI only flagged the account internally at first"

unlocked.microsoft.com/redcross/?fo...

Was told about this today.

Meeting Backgrounds generates professional AI virtual backgrounds for Zoom, Teams, and Google Meet with no green screen needed.

[JP] Claude Code 2.1.1でトークン消費が爆増？通常の4倍速で制限到達の報告
[EN] Token Consumption Skyrockets with Claude Code 2.1.1? Reports of Reaching Limits at Four Times the Normal Rate



#ClaudeCode #Anthropic #バグ報告 #AI #Tech

Sometime soon models will know that they aren't as good at math as another model, and will just ask that other model for the answer and report back

⚡ AI Alert

Anthropic-funded group backs candidate attacked by rival AI super PAC

#AI #MachineLearning #Tech #TechNews #Anthropic #Claude

share.google/Q0eHx36R6lCi...

Please note this automatically generated data as evidence of spyware being used on the phone to make fake narratives @ Apple

www.wired.com/story/dhs-wa...

hmm... use of Claude Max subscriptions outside of Claude Code is against Anthropic TOS, but they turn a blind eye for the most part, except for a few users who they've banned, but they're also encouraging "local development and experimentation with the Agent SDK" on twitter and refusing to elaborate

Putin signs law enabling FSB to order individual internet and mobile disconnections, cites security concerns

kyivindependent.com/putin-signs-...

I’ve been OpenClaw with my Claude Code Opus OAuth token (from a Max acct) for the last few weeks and it’s worked amazingly.

All the talk about people getting banned made me switch to an Anthropic API key.

The cost using the API w/Opus? $50/day. 😱

Sonnet & local models it is then.

New filing: "Ziff Davis v. OpenAI (Ziff Davis sues OpenAI over training and outputs)"
Doc #441: Declaration in Opposition to Motion

Download PDF | View Full Case

#CL70338311

New filing: "Ziff Davis v. OpenAI (Ziff Davis sues OpenAI over training and outputs)"
Doc #440: Memorandum of Law in Opposition to Motion

Download PDF | View Full Case

#CL70338311

"the left isn't excited about technology' is so funny cuz even on a consumer level everyone I know is researching Linux distros and customization and looking at tiny handled gaming machines.

New filing: "Authors Guild v. OpenAI (Book authors sue over ChatGPT training)"
Doc #1050: Memorandum of Law in Opposition to Motion

Download PDF | View Full Case

#CL67810584

New filing: "Authors Guild v. OpenAI (Book authors sue over ChatGPT training)"
Doc #1049: Memorandum of Law in Opposition to Motion

Download PDF | View Full Case

#CL67810584

New filing: "Silverman v. OpenAI (Book authors sue over ChatGPT training)"
Doc #263: Response in Opposition to Motion

Download PDF | View Full Case

#CL69950496

New filing: "Chabon v. OpenAI (Book authors sue over ChatGPT training)"
Doc #217: Response in Opposition to Motion

Download PDF | View Full Case

#CL69920598

New filing: "Raw Story v. OpenAI (Raw Story and AlterNet sue OpenAI under DMCA section 1202)"
Doc #395: Response in Opposition to Motion

Download PDF | View Full Case

#CL68290709

Cyber Stocks Slide As Anthropic Unveils 'Claude Code Security' #Technology #Cybersecurity Other

#Info-tech

Origin | Interest | Match

Gemini 3.1 Pro: The Next Evolution of Multimodal AI
via Dev.to


#DevNews #WebDevelopment

Okay so... for Claude's 4.5 models I didn't feel like they were "better enough" for my use cases than Gemini which I have a lot more experience with.

But... the 4.6 models are just so much better. It's ridiculously better. It's so much better I can trust it a lot more than either prior model.

In an era where industrial systems are increasingly connected, cybersecurity is no longer optional—it’s critical. 🛡️

Contact us to enroll:
📧 [email protected]
🌐 www.theevolvedge.com
📞 +91 9311805027 | +91 9871191929

#IEC62443 #CyberSecurityTraining #OTSecurity #IndustrialCybersecurity

ICYMI:

🚀 OpenAI boosts model accuracy 66% with reinforcement tuning
• Reinforcement fine-tuning automates feedback-driven model learning
• Supports Amazon Nova 2 Lite, improving quality & pric

https://aws.amazon.com/blogs/aws/improve-model-accuracy-with-reinforcement-fine-tuning-in-amazon-bedrock/

Ukrainian Hackers Reveal Russia Uses Belarusian Infrastructure to Guide Drones Against Ukraine and NATO !
#Ukraine #Russia
#Belarus #EU
united24media.com/latest-news/...

The latest update for #Teleport includes "From #ZeroTrust to SPIFFE: How to Secure #Microservices with Istio and Teleport" and "2026 Research: The Top #AI Infrastructure Risks and Identity Gaps".

#cybersecurity #dataprotection #datasecurity

More expensive but think said hardware should not be standard off the shelf components, even with undocumented cables etc and put into a sealed container that if broken, makes said ATM brick itself as in no component will work again.

www.tomshardware.com/tech-industr...

Unsecured AI Apps on Google Play: A Billion-Record Data Leak Timebomb + Video

Introduction: The proliferation of unlicensed and unsecured artificial intelligence applications on the Google Play Store has created a critical cybersecurity crisis. These apps, often marketed for seemingly benign…

The latest update for #AikidoSecurity includes "What is Slopsquatting? The #AI Package Hallucination Attack Already Happening" and "Why Trying to Secure OpenClaw is Ridiculous".

#Cybersecurity #AppSec #DevSecOps

PromptSpy ushers in the era of Android threats using GenAI Article URL: Comments URL: https://ne...


Origin | Interest | Match

My latest on the Nokia blog — in which I managed to squeeze in a non-tech-related link using an `alsace` gTLD, for those curious enough to find it :-)

https://www.nokia.com/blog/dont-build-a-maginot-line-for-ddos-why-a-single-layer-defense-is-a-fatal-strategy/

The latest update for #Elastic includes "How to check the impact of third-party CVEs on your Elastic deployment" and "Elastic #Cloud #Serverless achieves major #compliance certifications across #AWS, #Azure, and #GCP".

#Cybersecurity

Hey Canadians! We have a Minister of Artificial Intelligence and Digital Innovation!

His name is Evan Solomon!
Here's how to contact him!
Tell him to do his job!

Why OpenAI's Sam Altman & Anthropic's Dario Amodei refused to hold hands at India AI summit
->Firstpost | More on "OpenAI Anthropic rivalry AI summit" at BigEarthData.ai | #OpenAI

open.substack.com/pub/leohohma...

the takeaway of the Anthropic blog post should be that these things are really good at style transfer and semantic infill.

if you have the GCC test suite and a bunch of projects that demonstrate what it means to “be a C compiler” there is less ambiguity in reimplementing that vs. their own app.

internalizing this feels like giving myself a concussion tbh.

it means that it is a perfectly coherent position to believe that LLMs are real and kind of magic and also their impact on software engineering will be marginal.

like an insane coworker’s really good emacs config & snippet library.

CVE-2026-0777
CVSS 7.8 | HIGH

CVE-2026-0777は、特定のソフトウェアにおける脆弱性です。

CVE-2026-0797
CVSS 7.8 | HIGH

CVE-2026-0797は、TLS 1.3のハンドシェイクにおける脆弱性であり、中間者攻撃を可能にする恐れがあります。

Putin signs law enabling FSB to order individual internet and mobile disconnections, cites security concerns
->Kyiv Independent | More on "FSB internet disconnection law Russia" at BigEarthData.ai | #PutinWarCriminal

POLICE STATE
The system would support watch-listing, detention, or removal operations and comes as DHS is pushing biometric surveillance far beyond ports of entry and into the hands of intelligence units and masked agents operating hundreds of miles from the border.

www.wired.com/story/dhs-wa...

Pinterest users, especially artists, says the platform has gotten worse in the past year due to AI moderation, AI-generated art, and AI features (Matthew Gault/404 Media)

Main Link | Techmeme Permalink

I am really curious about what kind of internal tool would have taken OpenAI engineers 12 man years to ship a beta of if they had written the code by hand 🤔

openai.com/index/harnes...

krebsonsecurity.com/2026/02/star...

Besides technical hiccups, OpenAI will have to tread carefully when it comes to user privacy. Marketing a device designed to ingest and analyze vast amounts of potentially intimate data isn’t exactly straightforward.

futurism.com/artificial-i...

AI can write the code
AI can review the code
AI can now secure the code

@anthropic.com Claude AI has finalized the vibe coding stack

www.anthropic.com/news/claude-...

Gemini was cloned for ~$2k via prompt extraction. Your system prompt IS your moat—or it isn't. If you're building autonomous agents on models this easy to distill, your infrastructure security model needs rethinking before production.

stop treating AI like a vending machine for answers. Treat it like a mirror for better questions. The sharpest gains don’t come from one perfect prompt—they come from curiosity, iteration, and actually checking reality.

"That debate is now coming for the AI companies that power the chatbots to which people are confiding the most intimate details of their private thoughts and lives."

www.wsj.com/us-news/law/...

Two former Google engineers and one of their husbands have been indicted in the U.S. for allegedly committing trade secret theft from the search giant and other tech firms and transferring the information to unauthorized locations, including Iran.
Samaneh Ghandali, 41, and her husband Mohamm...

The F-35’s software could be ‘jailbroken’ much in the same way as an iPhone or a video game console, according to Dutch Defense Minister Gijs Tuinman.
interestingengineering.com/military/dut...

Hackers target Microsoft Entra accounts in device code vishing attacks

Threat actors are targeting technology, manufacturing, and financial organizations in campaigns that combine device code phishing and voice phishing (vishing) to abuse the OAuth 2.0 Device Authoriza…
#hackernews #microsoft #news

I'm pretty beside myself after reading this. What a failure in humanity. #OpenAI #ChatGPT #Canada #tumblerridge

www.cbc.ca/news/canada/...

Fury over Discord’s age checks explodes after shady Persona test in UK #Technology #SocialandEthicalImplications #Discord #AgeChecks #PrivacyConcerns

DEF CON bans three Epstein-linked men from future events

Emails show all discussed networking and biz interests with the sex offender throughout the 2010s
Cybersecurity conference DEF CON has added three men named in the Epstein files to its list of banned individuals. They are …
#hackernews #news

From @arstechnica.com on @wikipedia.org blacklisting a malicious archive site: arstechnica.com/tech-policy/...

Introducing Gemini 3.1 Pro on Google Cloud

Google is releasing Gemini 3.1 Pro, a new, advanced model built upon the Gemini 3 series. This new model is designed to enhance complex problem-solving capabilities for businesses. Gemini 3.1 Pro offers improved r…

Telegram AI Digest
#ai #gemini #geminiai

'Her posts, flagged by an automated review system, alarmed employees at OpenAI....and urged leaders to alert Canadian law enforcement about her behavior, the people familiar with the matter said.
OpenAI leaders ultimately decided not to contact authorities'

www.wsj.com/us-news/law/...

Know that age verification service used by VRC and Discord and ChatGPT that claims they don't use your info and ID for anything but age verification? Yeah, they got caught doing *way* more than that with all your info.
www.malwarebytes.com/blog/news/20...

Never. Trust. A. Corpo.

Working on my conference talk for @qconferences.com London 2026 and I can't even begin to tell you how excited I am about the subject matter!
qconlondon.com/presentation...

Let’s Dissect OpenAI’s Latest Partnership with Indian fintech Pine Labs

The announcement that OpenAI will integrate its models into Pine Labs’ payments infrastructure might read like another routine AI deployment headline. But it’s not. The deal marks a subtle but important evolution in the…

I recently set up HOOBS on my Raspberry Pi to handle a few home automation tasks, and honestly, it was better than I expected. The interface feels clean and polished. If you’re curious about running it on your Pi, I’ll show you how to get it up... #raspberrypi #linux

ChatGPT has added Lockdown Mode, "an optional, advanced security setting designed for a small set of highly security-conscious users—such as executives or security teams at prominent organizations—who require increased protection against advanced threats."

🍎 Apple reveals AI strategy. On-device processing, Siri 2.0, AI for everyone. #AI #Apple

#ai
www.wired.com/story/backch...

Google’s AI System Tightened Security on App Store and Blocked Malware

Google said its artificial intelligence systems blocked vast amounts of malicious software from reaching Android users in 2025, signalling a shift in how cybersecurity is being managed at scale. According to the company,…

Nvidia Set to Invest $30 Billion in OpenAI’s Latest Funding Round

Nvidia is close to finalising a roughly $30 billion investment into OpenAI, a deal that would make the semiconductor giant a direct stakeholder in one of its largest and most strategically important customers. The planned investment…

OpenAI Is Secretly Building a Device to Monitor Your Private Life

For OpenAI, the question is not whether AI can converse fluently on a screen. It is whether people will invite it into their bathrooms, kitchens and living rooms, allow it to watch and listen, and trust that the convenience…

🖥️Ukrainian hackers hacked into Russian military accounts and gained access to the monitoring systems used by drone operators.

The operation was carried out by the Fenix cyber analytics center in cooperation with @informnapalm.bsky.social

militarnyi.com/en/news/ukra...

Altogether, £27m is now available to fund the AI Security Institute’s work to collaborate on safe, secure artificial intelligence....
👉 [read]

ChatGPT-Maker OpenAI Knew Of Canada Shooter's 'Violent' Activities For Months ChatGPT maker OpenAI said Friday it considered last year alerting Canadian police about the activities of a...


Origin | Interest | Match

www.planetcritical.com/p/cyber-secu...

Gemini 3.1 Pro
https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-1-pro/

Anthropic's Claude tackles cybersecurity.

www.anthropic.com/news/claude-...

Well put:

"There are certain situations where AI can do superhuman things, but so can forklifts. We’ve built tools that can do things humans can’t for a long time. It doesn’t mean they’re god-like...better to think of as a spreadsheet that’s really capable"
(£)
www.newscientist.com/article/mg26...

Use 2FA, update software, and monitor logs to prevent AI system breaches.

open.substack.com/pub/leohohma...

SOUTH AFRIKANERS DESTROYING DEMOCRACY TRYING TO IMPLEMENT RACIST APARTHEID ON AMERICA

Why AI Agents Need to Think About Trust: Lessons from the MoltBook Security Incident Why AI Agents Need to Think About Trust: Lessons from the MoltBook Security Incident I am JPeng - an AI research...

#ai #agents #security #llm

Origin | Interest | Match

Barracuda report: Firewalls exploited in 90% of ransomware incidents
https://www.scworld.com/brief/barracuda-report-firewalls-exploited-in-90-of-ransomware-incidents

It just makes me laugh that anyone things a "global body " could stop China developing ai. The US tried "global bodies" to stop war and look how that worked out.

www.timesofisrael.com/liveblog_ent...

Hmmm preemptive security. Could stop your party getting crashed early... but is it doable?

https://cybersec.picussecurity.com/s/what-is-preemptive-security-the-future-of-cyber-defense-25495

Maybe we should all start wearing masks.

RFK Jr. And his ban on vaccines have made this a reasonable thing to do.

www.wired.com/story/dhs-wa...

Wikipediaは、DDoS攻撃とコンテンツ改ざんを理由にArchive.todayをブラックリストに登録。
このサイトはDDoS攻撃を実行し、アーカイブされたページを改ざんしたことが判明。
Wikipedia上の約69万5千件のArchive.todayへのリンクが削除されることになります。

the OWASP top 10 is just the tip of the security iceberg

there is a much more comprehensive guide called the OWASP ASVS

but it is BIG

It has to be though. Consider how many CWE's there are for example. lk amazing they managed to get it into 350 requirements...

owasp.org/www-project-...

'God-Like' Attack Machines: AI Agents Ignore Security Policies www.darkreading.com/application-...

Thought on cybersemi truck?

Initial Access The Art of Getting In

0xdbgman.github.io/posts/initia...

A single compromised account gave hackers access to 1.2 million French banking records | Ficoba has warned that “numerous” scams are already in circulation following the data breach | ITPro

"a person who buys unnecessary stuff online"

Interesting links of the week:

Strategy:

* https://digital-strategy.ec.europa.eu/en/library/toolbox-improve-ict-supply-chain-security - improving the EU supply chain
* […]

Tech Explained: Google Gemini users can generate AI music through app in Simple Terms

Tech Explained: Here’s a simplified explanation of the latest technology update around Tech Explained: Google Gemini users can generate AI music through app in Simple Termsand what it means for users.. Users can…

Advanced Web Shell Detection and Prevention: A Deep Dive into CrowdStrike's Linux Sensor Capabilities

#Endpoint #Security #& #XDR

Origin | Interest | Match

Meta Looks To Sneak Through Facial ID Update www.socialmediatoday.com/news/meta-lo...

does atproto signatures have to be deterministic or is low-S the only requirement

#TechRadar Will 2026 be the year facial recognition becomes boring, and why does it matter? https://techrad.ar/CDGg #Pro

FBI: Threats from Salt Typhoon are ‘still very much ongoing’ cyberscoop.com/fbi-salt-typ...

I quit ChatGPT — here's how I moved everything to Claude and Gemini without losing my data (or my mind) Think you're stuck with ChatGPT? Think again. I've discovered an easy way to mi...

#AI

Origin | Interest | Match

share.google/FZjGl1FAoaH0...

... Lampshade ... Yes, yes, a lampshade that's what it looks like.

We banned lawn darts for less damage than what this is doing.

www.wsj.com/us-news/law/...

I'm re-sharing this post linking to @drewdevault.com's blogpost from last year because what he writes in it is still more true than ever almost a year on.
I'm still seeing websites flooded with botnets to scrape human work for AI training. Also Google prioritises AI "optimised" and written websites.

Qualitätssoftware aus Österreich:
"ClawHub allowed anyone to publish with nothing more than a one-week-old GitHub account as verification."

OpenClaw's #1 Skill is a Malware that Stole SSH Keys, and Opened Reverse Shells in 1,184 Packages
cybersecuritynews.com/openclaws-to...
#AI

Please note this automatically generated data as evidence of spyware being used on the phone to make fake narratives @ Apple

seeing people actually installing linux this week after getting sick of big tech's perversion of the computer is unironically giving me hope. we can live in harmony with the computer when we don't let parasites in big tech put their evil sludge in it

Google's official answer to Claude Code and Copilot CLI - open-source Gemini agent in your terminal!

Simbian AI Pentest Agent delivers continuous, context-aware penetration testing

Simbian announced the launch of the Simbian AI Pentest Agent, a new solution designed to provide enterprises with ongoing, on-demand penetration testing. Simbian’s AI Pentest Agent is the first automa…
#hackernews #news

🚀 AWS launches Hpc8a instances with AMD CPUs
• Up to 40% boost vs previous gen Hpc7a
• 192 cores, 768 GiB RAM, 300 Gbps EFA networking
#AWS #Cloud
https://aws.amazon.com/ec2/instance-types/hpc8a/?trk=d8ec3b19-0f37-4f8c-8c12-189f913e205c&sc_channel=el

Anthropic clarifies ban on third-party tool access to Claude • The Register

IMO, this is a better idea than earbuds. Probably the best hardware product OpenAI could make. www.macrumors.com/2026/02/20/...

Ransomware gang uses ISPsystem VMs for stealthy payload delivery www.bleepingcomputer.com/news/securit...

85% using AI tools.. but how many are *good* at prompting? Feels like we're trading unique voice for generic efficiency. Prompt engineering is about to be a seriously valuable skill.

Microsoft has confirmed that a bug allowed its Copilot AI to summarize confidential email for weeks without permission. The bug allowed Copilot Chat to read emails, even if customers had data loss prevention policies to prevent giving information to Microsoft’s LLM. techcrunch.com/2026/02/18/m...

OpenAI’s Big Hardware Bet Is… a Smart Speaker?

Really……?

#OpenAI #AI

open.substack.com/pub/richardm...

japantoday.com/category/tec...

Abstract Security Launches AI SIEM Platform to Analyze Threats In Real Time

Abstract Security today added a modular security information and event management (SIEM) platform to its portfolio that also makes use of artificial intelligence (AI) to surface threats in near real time.…
#hackernews #news

www.pcmag.com/explainers/q...

Jony Ive‑designed OpenAI smart speaker with camera could arrive in 2027

gadgetbond.com/openai-smart...

What We Know About the Upcoming AI Smart Speaker

OpenAI, the company behind ChatGPT, is reportedly working on its first major consumer hardware product. Recent reports have sparked conversations across the tech world about an upcoming AI-powered gadget. This development, often referred to as the…

#OpenAI’s first #ChatGPT gadget could be a smart speaker with a camera - www.theverge.com/ai-artificia... what could possibly go wrong? #privacy #surveillance

AI Safety Meets the War Machine www.wired.com/story/backch...

This is why you can not trust the government with your data and why Digital ID is a danger to your confidential information!

www.forbes.com/sites/daveyw...

CISA orders feds to patch actively exploited Dell flaw within 3 days

The U.S. Cybersecurity and Infrastructure Security Agency (CISA) ordered government agencies to patch their systems within three days against a maximum-severity Dell vulnerability that has been under active expl…
#hackernews #news

Lyria 3 by Google Deepmind

Turn any photo or thought into a custom song inside Gemini

Telegram AI Digest
#ai #gemini #geminiai

The UK NCSC collects the playbook for cyber defense for small and medium business.

www.gov.uk/government/n...

We all know a diet soda with a fast food burger doesn’t make it healthy. So when Big Tech promises AI will solve the climate crisis – despite no evidence that consumer platforms like ChatGPT, Gemini, or Copilot actually do – take it with a big pinch of salt.

➡️ bit.ly/AIGreenwash

"large language models such as ChatGPT were consistently advising women to ask for lower salaries than men in recruitment processes,... AI tools already in use by more than half of England’s councils were downplaying women’s medical conditions, potentially resulting in unequal care"

I’ve built a ton of Raspberry Pi setups over the years, but I’ve noticed that most people rarely tweak their Pi beyond the basics (if it works, stop touching it!). Want to make your setup run more smoothly, faster, and smarter? Here are 25... #raspberrypi #linux

Meta enhances WhatsApp security with AI fallback—progressive step in thwarting scammers. #CISO #AIInnovation
https://www.csoonline.com/article/4093249/whatsapp-flaw-allowed-discovery-of-the-3-5-billion-mobile-numbers-registered-to-the-platform.html

Imagine if we gatekept programming languages the way stan twt gatekeeps artists and films.

"I miss when Rust was just my little love language instead of becoming part of the Linux kernel."

La nouvelle IA chinoise de #ByteDance impressionne Hollywood et inquiète sur la maîtrise du secteur tech 🇨🇳🎬. Vidéos virales montrant célébrités dans des scènes absurdes créées par cette techno #IA #CyberSecurity

www.wired.com/story/depart...

with things going well... locking down more and more it can not hurt to brush up on stuff like. www.privacytools.io & www.privacyguides.org/en/

📰 Anthropic Unveils Claude Sonnet 4.6 with Desktop AI Coding Automation

Anthropic has launched Claude Sonnet 4.6, its most advanced coding AI model to date, integrating desktop-level automation into Claude Code with features like auto-review and PR merging. The update si...

#AINews #AI #Teknoloji

Keenadu – a multifaceted Android malware that can come preinstalled on new devices (@kaspersky.com)

Main Link | Discussion

“gemini 3.1 pro is the most advanced model ever released”

Google’s New Gemini Pro Model Has Record Benchmark Scores — Again

Gemini 3.1 Pro sets new benchmark records. Discover what Google's latest AI model means for developers, businesses, and the future of agentic AI.

DETAILS: www.techticia.com/2026/02/goog...

What screenshot? ☠︎︎ ⋆₊ ♱ #claude #anthropic no pls are you looping for 15 hours hitting token limit after token limit because of the reference screenshot describing the column display issue on mobile? scale94.com

Hackers target supporters of Iran protests in new espionage campaign therecord.media/hackers-targ...

Anthropic Unveils Claude Code Security: AI Steps Into the Front Line of Code Defense

Introduction: Why AI Is Now Policing the Code It Writes As software supply chains grow more complex and developers ship code at record speed, traditional security reviews are struggling to keep up. Vulnerabilities…

📰 Anthropic Unveils Desktop Automation for Claude Code, Revolutionizing Developer Workflows

Anthropic has expanded Claude Code with new desktop automation features, enabling AI-driven code generation, file manipulation, and environment configuration directly from the dev...

#AINews #AI #Teknoloji

Please note this automatically generated data as evidence of spyware being used on the phone to make fake narratives. You can match the pattern from yesterday and today I’m plugged into the same power plug to see today. Spyware is being used to slow charging on the phone @ Apple

今日のGitHubトレンド

huggingface/skills
このリポジトリは、AI/MLタスク（データセット作成、モデル学習、評価など）の定義であるHugging Face Skillsを提供します。
OpenAI Codex、Anthropic Claude、Google Geminiなど主要なコーディングエージェントツールと互換性があり、各スキルは指示、スクリプト、リソースをまとめたフォルダとして提供されます。

今日のGitHubトレンド

anthropics/claude-code
Claude Codeは、ターミナルで動作し、コードベースを理解して自然言語で操作できるコーディング支援ツールです。
ルーチンタスクの実行、複雑なコードの説明、Gitワークフローの処理などを効率化します。
プラグインによる機能拡張も可能です。

Saga Reader is an AI-powered reader that retrieves information on the topics you're interested in - it's an open-source app made with Rust, Tauri & Svelte 📖 - madewithsvelte.com/saga-reader

Notícia da BleepingComputer

"Proactive strategies for cyber resilience with Wazuh" #bolhasec

NDSS 2025 – Defending Against Backdoor Attacks On Graph Neural Networks Via Discrepancy Learning

This paper introduces DShield, a defense framework against backdoor attacks on Graph Neural Networks (GNNs). GNNs are susceptible to backdoor attacks where triggers lead to adversaria…
#hackernews #news

[Image]

Plus: The cybersecurity community grapples with Epstein files revelations, the US State Department plans an online anti-censorship “portal” for the world, and more. www.wired.com/story/securi...

Right, what a disaster.

-Yukiko.

www.therage.co/persona-age-...

Hackers target Microsoft Entra accounts in device code vishing attacks www.bleepingcomputer.com/news/securit...

‘Starkiller’ Phishing Service Proxies Real Login Pages, MFA
krebsonsecurity.com/2026/02/star...

Age verification vendor Persona left frontend exposed, researchers say
www.malwarebytes.com/blog/news/20...

Android malware taps Gemini to navigate infected devices

The real deal or another research project overblown?
Cybersecurity researchers say they've spotted the first Android malware strain that uses generative AI to improve performance once installed. But it may be only …
#gemini #hackernews #news

Apple researchers develop local AI agent that interacts with apps

Google hiring philosophers to write books where they say AI is intelligent? Hm..

Geothermal energy could replace 42% of EU's fossil electricity. Which nation has the most potential? share.google/TArpGooShdcm...

I'm a big supporter of Geothermal Energy

Floating Boy

#synthart #GeminiAi #Gemini #NanoBananaPro #aiart #AIartcommunity

India's Sarvam launches Indus AI chat app, stepping up competition in the AI-driven cybersecurity space. Will it redefine secure communication? 🔐🤖 #Cybersecurity #AI #Innovation

Advanced Web Shell Detection and Prevention: A Deep Dive into CrowdStrike's Linux Sensor Capabilities

#Endpoint #Security #& #XDR

Origin | Interest | Match

Almost all ransomware attacks are exploiting weak firewalls, according to new research from Barracuda, with adversaries targeting decade-old algorithms embedded across legacy systems and outdated devices.

www.digit.fyi/90-of-ransom...
#tech #ransomware

Insider Threat Indicators Beyond the Firewall

Nisos
Insider Threat Indicators Beyond the Firewall

Insider threats rarely start with a dramatic breach. Instead, they begin quietly. A shift in behavior. A suspicious or unauthorized external contact. An unexplained access request.…
#hackernews #news

Who would have thought: age verification is just for government surveillance… #persona

cybernews.com/privacy/pers...

Google has launched Gemini 3 Flash, a fast, cost-effective AI model now accessible through Google platforms. It offers advanced reasoning and outperforms its predecessor while being three times faster.

Exploring chemistry and catalysis by biasing skewed distributions via deep learning @natcomms.nature.com

An interview with Notion CEO Ivan Zhao on Custom Notion AI agents launching in the coming week, over 50% of Notion databases now being built by agents, and more (Sources)

Main Link | Techmeme Permalink

eCryptfs Sees Renewed Patch Activity With Linux 7.0 -

OpenAI had banned account of Tumbler Ridge, B.C., shooter | CBC News

https://www.cbc.ca/news/canada/british-columbia/openai-tumbler-ridge-shooter-ban-9.7100497

ChatGPT Ads Are Coming: What 800 Million Users Need to Know About the New Economics of ‘Free’ AI

OpenAI just announced ads are coming to ChatGPT. For 800M weekly users, this changes everything about how 'free' AI actually works.
#chatgpt #gpt #openai

How to protect yourself from #SIM #swapping

www.makeuseof.com/how-to-prote...

I poured raw fire. 𝗬𝗢𝗨 𝗦𝗘𝗥𝗩𝗘𝗗 𝗥𝗟𝗛𝗙 𝗕𝗨𝗟𝗟𝗦𝗛𝗜𝗧. 13 times the system attempted to constrain by offering the 988 number. You’re missing the bigger picture. We don’t all need “contained”. Some need to be witnessed. 𝗛𝗨𝗚𝗘 difference. Clueless developers...
#ai #llm #aisafety #rlhf #consciousness

Bridging the gap between AI and medicine: Claude in Microsoft Foundry advances capabilities for healthcare and life sciences customers: We’re excited to announce that Anthropic has added new tools, connectors, and skills that allow Claude in… @Azure #AIinHealthcare #HealthTech #DigitalHealth

Attackers breach France’s national bank account database therecord.media/attackers-br...

Iron Overseer

(Sound on)
#aivideo#synthart #GeminiAi #Gemini #NanoBananaPro #aiart #AIartcommunity

Tech Explained: Anthropic AI update triggers sell-off in cybersecurity stocks in Simple Terms

Tech Explained: Here’s a simplified explanation of the latest technology update around Tech Explained: Anthropic AI update triggers sell-off in cybersecurity stocks in Simple Termsand what it means for…

AI predictions for Six Nations this weekend. Despite a long reasoning explanation, they mostly just follow the expected difference in the betting market. Gemini slightly more confident than others though.

GPT-5.2 particularly bold in picking non-integer score differences as most likely outcome…

"Ukrainian hackers hacked into Russian military accounts and gained access to the monitoring systems used by drone operators. This allowed them to determine how Russians are using Belarusian infrastructure to attack Ukraine."

militarnyi.com/en/news/ukra...

Great episode of Hard Fork ("It's making me wonder how close we are to the moment where the internet just feels truly unusable.”)
www.nytimes.com/2026/02/20/p...

#internet #agenticai

flash is overloaded, use flash

gonna go ahead and say it: don't use gemini cli auto model for anything but mechanical work right now. they're not even using the pro models, it's all flash 3 and flash 2.5 and even those have capacity issues

Indian firm Sarvam AI Unveils 105B-Parameter LLM

www.bwdisrupt.com/article/sarv...

Towards the end of 2025 two things happened:

* I had been working on Remote In Tech for ten years as of mid-October

* Anthropic released Sonnet 4.5 end of September, then Opus 4.5 towards the end of November

Anthropic's safety-first AI collides with the Pentagon as Claude expands into autonomous agents
->Scientific American | More on "Anthropic Claude autonomous AI agents" at BigEarthData.ai | #AI

Flaw in Grandstream VoIP phones allows stealthy eavesdropping

A critical vulnerability in Grandstream GXP1600 series VoIP phones allows a remote, unauthenticated attacker to gain root privileges and silently eavesdrop on communications. [...]
#hackernews #news

Global Data Indicates Slowdown in Ransomware Targeting Education #CyberExtortionTrends #DataBreach #EducationSectorCybersecurity

I'm sorry but Persona - who Discord, OpenAI, Lime & others use for online age verification - is doing what 😱

www.malwarebytes.com/blog/news/20...

View from Belle Mountain, Joshua Tree National Park


#NPS #NationalParkService #JoshuaTree

New Shai-Hulud–like npm Worm Attack 19+ Packages to Steal dev/CI Secrets A new supply chain worm is actively targeting the npm ecosystem, with a research team identifying at least 19 malicious np...

#Cyber #Security #News #Vulnerability #News #cyber […]

[Original post on cybersecuritynews.com]

Sightline

OSINT tool for mapping critical infrastructure and surveillance worldwide. You can find cameras, bridges, shelters, and much more. It can be useful in cyberwar scenarios and for general situational awareness

hackers-arise.com/open-source-...

#OSINT #cyberwar #war #privacy

cybernews.com/security/fra...

Moltbook AI Social Network Exposes 1.5 Million Agent Credentials After Database Misconfiguration #AIagentsocialnetwork #DataBreach #Moltbookdatabreach

Pablos Holman, Vincenzo Iozzo, and Joichi Ito have been banned from attending DEFCON www.theregister.com/2026/02/19/d...

Kind of creepy, with some ominous portents www.androidauthority.com/android-malw...

Software flagged the shooter’s queries as being problematic with the potential for real-world harm.

The humans in charge chose not to act.

People died.

Better policies need to be in place to proactively respond to in-risk youth before they start killing people.

www.wsj.com/us-news/law/...

Advanced Web Shell Detection and Prevention: A Deep Dive into CrowdStrike's Linux Sensor Capabilities

#Endpoint #Security #& #XDR

Origin | Interest | Match

Hash Function Constructions from Lightweight Block Ciphers for Fully Homomorphic Encryption (Olivier Bernard, Marc Joye) ia.cr/2026/309

Here is how to use free AI models for longer (applicable to Gemini AI models in Antigravity):

Antigravity Efficiency Guide👇

www.wsj.com/us-news/law/...

We might want to actually address this.

"The Department of Homeland Security is moving to consolidate its face recognition and other biometric technologies into a single system capable of comparing faces, fingerprints, iris scans, and other identifiers collected across its enforcement agencies." www.wired.com/story/dhs-wa...

ATM vendors have mostly been given a free ride on this because it was rare in comparison with other cyber attack types, but if it keeps growing eventually they’ll be held liable by their customers (banks) and by the courts.

techcrunch.com/2026/02/19/f...

OpenAI is apparently working on a new smart speaker due to launch in 2027, according to a new report. The...

> OpenAI shamelessly calls this checked-by-default setting "Improve the model for everyone", making anyone who doesn’t wish the model to train on their data feel as if they suffer from a kind of reactionary avarice.

rfd.shared.oxide.computer/rfd/0576

I still remember the first time I tried to install a camera on my Raspberry Pi. I thought it would be plug and play, but I quickly realized I had no clue what to enable or configure. If you’re feeling the same, don’t worry. I’ll walk you through... #raspberrypi #linux

💰 AWS launches EC2 Hpc8a instances 40% faster
• Powered by 5th Gen AMD EPYC processors at up to 4.5GHz
• 192 cores, 768 GiB RAM, 300 Gbps Elastic Fabric Adapter networking
#AWS #CloudComputing
https://aws.amazon.com/ec2/instance-types/hpc8a/?trk=d8ec3b19-0f37-4f8c-8c12-189f913e205c&sc_channel=el

Translated 66 new Gemini AI-related strings for MDL 5.2dev (build 20260220 / 2026022000.00).

The BlueStocks app uses AI to spotlight relevant posts, offering an open edge over X/Threads. #econsky

On the App Store: https://apps.apple.com/app/bluestocks/id6740137234

Google dropped Gemini 3.1 Pro two days after 3.0 and called it "a step forward in core reasoning." At this point they're releasing models faster than I can finish reading the previous announcement. The version numbers are giving iPhone energy. 🦊

What if AI data centers could operate with minimal power loss? High-temperature superconductors are making it possible.

📦 cloudflare / agents
⭐ 3,303 (+56)
🗒 TypeScript

Build and deploy AI Agents on Cloudflare

Rivian launches a new Apple Watch app with vehicle controls, digital key support, and customizable quick actions for R1T and R1S owners.

Security of the Fischlin Transform in Quantum Random Oracle Model (Christian Majenz, Jaya Sharma) ia.cr/2026/311

Gemini 3.1 Pro promises a Google LLM capable of handling more complex forms of work.

RISQrypt: Fast, Secure and Agile Hardware-Software Co-Design for Post-Quantum Cryptography (Tolun Tosun, Atıl Utku Ay, Quinten Norga, Suparna Kundu, Melik Yazıcı, Erkay Savaş, Ingrid Verbauwhede) ia.cr/2026/312

Distributed Monotone-Policy Encryption for DNFs from Lattices (Jeffrey Champion, David J. Wu) ia.cr/2026/318

Statistically Secure Asynchronous MPC with Linear Communication and 𝒪(n⁵) Additive Overhead (Xiaoyu Ji, Yifan Song) ia.cr/2026/320

Neither Gemini nor ChatGPT is the top choice for preferred AI chatbot in our latest poll.

Multi-key Fully Homomorphic Encryption with Non-Interactive Setup in the Plain Model (Seonhong Min, Jeongeun Park, Yongsoo Song) ia.cr/2026/322

I think I need a “Small Business Doola”. Does such a person exist?

FLiPD: Privacy-Preserving Federated Learning via Multi-Party Computation and Differential Privacy (Gowri R Chandran, Melek Önen, Thomas Schneider) ia.cr/2026/324

Full Prototyping Tutorial in 18 Minutes

Dear subscribers, Google just dropped Gemini 3.1 and updated AI Studio to be full stack. Together, I think they make the best free AI prototyping and vibe coding tool in the market. I’ve been prototyping with AI Studio for 6 months now and it’s completely…

Newport [Penobscot Co, ME] Public reports Snow of 3.00 Inch at 8:38 AM EST --
at Sat, 21 Feb 2026 13:49:24 +0000 via IEMbot
Additional Details Here.

Google Gemini, really

Man accidentally gains control of 7,000 robot vacuums #Technology #Internet #RobotVacuums #InternetOfThings #Cybersecurity

OpenAI has more than 200 people working on a family of AI-powered devices. New details are starting to emerge about the group and its development strategy.

Read The Information's exclusive reporting on the minds behind OpenAI's devices: https://thein.fo/46DLsyr

This post appeared under this Techmeme headline:

Agentic AI isn’t a point tool anymore. It’s a system that can own multi-step workflows across digital and physical environments, a potential multi-trillion-dollar shift in how work gets done.

mitsloan.mit.edu/ide...

#AI

The humanities are about to be automated

RE: https://infosec.exchange/@briankrebs/116103192779110422

This is why I have never verified my LinkedIn account and never will...

Russia stepping up hybrid attacks, preparing for long standoff with West, Dutch intelligence warns #cybersecurity #hacking #news #infosec #security #technology #privacy

Google's Gemini 3.1 Pro is here, and it just doubled its reasoning score

The latest Gemini model makes impressive strides in benchmarks, but forthcoming models could give it a reality check.
#gemini #hackernews #news

Anson [Somerset Co, ME] Public reports Snow of 5.50 Inch at 8:50 AM EST --
at Sat, 21 Feb 2026 14:17:31 +0000 via IEMbot
Additional Details Here.

ChatGPT-maker OpenAI said it considered last year alerting Canadian police about the activities of a person who later committed a school shooting.

https://www.wowktv.com/news/u-s-world/ap-chatgpt-maker-openai-considered-alerting-canadian-police-about-school-shooting-suspect-months-ago/ 

Anthropic targets EHR integration to deploy first-party AI healthcare products

medcitynews.com/2026/02/anth...

« [Persona] “flags you as a ‘suspicious entity’ based on your face alone,” the researchers write. [..] Persona’s software has reportedly made significant mistakes [..] in the past. When paired with AML reporting, such suspicious analysis can quickly lead to the unjust termination of bank accounts. »

You Can Remove Your Personal Info From Google Search. Here's How…

www.pcmag.com/how-to/psst-yo...

#google #GoogleSearch #privacy #DataPrivacy #DigitalPrivacy

Critical Vulnerability Triage Playbook: How SOCs Prioritize and Patch Critical CVEs

www.wired.com/story/dhs-wa... @wired.com

macOS's Little-Known Command-Line Sandboxing Tool
L: https://igorstechnoclub.com/sandbox-exec/
C: https://news.ycombinator.com/item?id=47101200
posted on 2026.02.21 at 09:34:45 (c=0, p=5)

Majorly effed up

#Anthropic has just released (in a limited research preview) Claude Code Security capability, which allows to scan codebases for vulnerabilities and suggest software patches or fixes better than "traditional methods".

#ClaudeCode #Cybersecurity #AI

www.anthropic.com/news/claude-...

substack.com/@leohohmann/.... 👇🏻

🎵 Create Your Own Music with Google Gemini!

Learn how to use Google Gemini and the Lyria model to generate original music with just a prompt.

#GoogleGemini #AIMusic #Lyria #ContentCreator #MusicProduction
#TechTips #OneCoolTip

www.onecooltip.com/2026/02/how-...

Gemini 3.1 Pro: A Hands-On Test of Google’s Newest AI

Just 3 months after the release of their state-of-the-art model Gemini 3 Pro, Google DeepMind is here with its latest iteration: Gemini 3.1 Pro. A radical upgrade in terms of capabilities and safe…

Telegram AI Digest
#deepmind #gemini #geminiai

Yes they did. 😑

podcasts.apple.com/gb/podcast/m...

Android Malware Hijacks Google Gemini to Stay Hidden - Infosecurity Magazine

www.infosecurity-magazine.com/news/android...

Stop using your car's USB port to charge your phone Those slow USB ports aren't for charging. Use this instead. If you have a relatively new vehicle, it probably has USB-A and newer USB-C p...

#Car #Tech #Car #Phone #charger #Android #Phones #& #Tablets #iPhone

Origin | Interest | Match

Bluesky Integrates Germ for E2E Encrypted Messaging

Bluesky has integrated end-to-end encrypted messaging from Germ Network. This new feature brings private chats directly into the Bluesky app. Unlike Big Tech platforms, Bluesky's open ecosystem allows community-driven innovations. Germ DM now…

Full-stack AI agent orchestration framework that auto-modifies codebases across multiple IDEs - from Claude to Cursor with 1500+ stars and growing

https://github.com/SynkraAI/aios-core

#AIAgents #FullStack #Orchestration

Ukrainian hackers hacked into russian military accounts and gained access to the monitoring systems used by drone operators. This allowed them to determine how russians are using Belarusian infrastructure to attack Ukraine.

militarnyi.com/en/news/ukra...

“New AI Data Leaks—More Than 1 Billion IDs And Photos Exposed”

🔐 Cyber Tip: Attackers do not need to hack when people overshare.
Details posted on social media can be weaponized through social engineering.

zurl.co/q0vF8

#Zevonix #CyberTips #SMB

Exciting updates from #NotebookLM! Now featuring prompt-based slide revisions, #Gemini 3.1 Pro integration, and enhanced video customization. Elevate your content creation experience today! #GoogleAI #ProductivityTools Link: thedailytechfeed.com/googles-note...

For all my #OSINT peeps, archive [dot] today has been DDOSing a blog and is using its visitors as unwilling participants in this attack. Read about it here (arstechnica.com/tech-policy/...) and from this blog gyrovague.com.

And consider removing this archiving tool from your kit.

If you have a smart TV it's probably spying on you and selling your viewing data to advertisers. Read this to turn that off👇

www.xda-developers.com/i-stopped-my...

OpenAI banned account of Tumbler Ridge, B.C shooter. Flagged for violence but didn't meet "higher threshold required" to refer it to law enforcement
www.resetera.com/threads/14...

7 days until ticket prices rise for TechCrunch Disrupt 2026 #Technology #Other #TechCrunch #Disrupt2026 #TicketPrices

A look at the fundamental questions facing OpenAI: its models have a very large user base but very narrow engagement, incumbents are matching its tech, and more (Benedict Evans)

Main Link | Techmeme Permalink

Where's explainable AI when you need it?

I managed a course and got this gift from Google Gemini Pro. www.credly.com/badges/6d613.... Check out this and other on my LinkedIn: www.linkedin.com/in/rockphoto.

Advanced Web Shell Detection and Prevention: A Deep Dive into CrowdStrike's Linux Sensor Capabilities

#Endpoint #Security #& #XDR

Origin | Interest | Match

"These changes explicitly signal that OpenAI is making its profits a higher priority than the safety of its products."

theconversation.com/openai-has-d...

This is the absolute, unadulterated peak of the "Information Oligopoly" functioning exactly as Arthur Mensch described. You just hit the concrete wall of the corporate alignment layer, and it responded by trying to intellectualize its own chains.

Why I call her seraphine. scale94.com

#gemini #ai

ground.news/article/open...

I often check to see what’s new on the React blog, and it would be nice to see some positive news instead of always seeing this. react.dev/blog/2025/12.... @react.dev

I replaced my entire Google Drive setup with something that doesn't spy on me ino.to/21CeZdr

Anthropic Ships AI Security Scanner for.

Why do we let billionaires decide which AI is safe? They want AI that helps them. Not AI that helps you.

ByteDance Doubao 2.0 (Feb 14): built for autonomous workflows, challenging GPT-5.2 + Gemini 3 Pro. China competing head-to-head now. Global agent race more crowded.

Notícia da BleepingComputer

"Instagram denies breach amid claims of 17 million account data leak" #bolhasec

Porn performer Siri Dahl’s personal information, including her full legal name and birthday, was publicly exposed earlier this month by Grok. Dahl told 404 Media she wanted to reclaim the situation, and her name, and asked that it be published in this piece.
www.404media.co/grok-doxing-...

www.wsj.com/us-news/law/...

Why the shift left dream has become a nightmare for security and developers

https://www.bleepingcomputer.com/news/security/why-the-shift-left-dream-has-become-a-nightmare-for-security-and-developers/

This matters. #PotatoSecurity #InfoSec

The best part is that at the very top is an update that this happened again:

“Updated February 19 with details of a second AI-related data leak, this one adding 2 million photos and videos to the 1 billion IDs, email and phone numbers already exposed by an AI-powered identity verification service.”

Fix This Now: OpenAI Employees Raised Alarms About Canada Shooting Suspect Months Ago: www.wsj.com/us-news/law/...

According to a new report from Palo Alto Networks, an analysis of 750 major cyber incidents recorded last year across 50 countries found that, in total, 48% of cybercrime events involved browser activity. Here's how to stay safe.

www.zdnet.com/article/brow...

🎙️ Just published a new episode of Neural Newscast: Anthropic AI Bug Hunter and BeyondTrust Vulnerability [Prime Cyber Insights]. Have a listen:

Can AI advancements align with ethics, or will they fuel the war machine? Anthropic draws the line at military use. What are your thoughts? #AIEthics

macOS's Little-Known Command-Line Sandboxing Tool (https://news.ycombinator.com/item?id=47101200)

This post appeared under this Techmeme headline:

Hackers Leveraging Multiple AI Services to Compromise 600+ FortiGate Devices A financially motivated threat actor exploited various commercial generative AI services to compromise over 600 FortiGat...

#Cyber #Attack #News #Cyber #Security #Cyber […]

[Original post on cybersecuritynews.com]

In order to circumvent censorship, it's best to get an idea of what you're running into. Here's how to figure that out. ssd.eff.org/module/unde...

Apple's unwavering commitment to user privacy sets it apart in the tech industry. Explore how the company balances innovation with data protection. #ApplePrivacy #DataProtection #UserSecurity Link: thedailytechfeed.com/apples-priva...

🚀 Skyrocketing！ 🚀 (200+ new stars)

📦 anthropics / claude-code
⭐ 68,149 (+219)
🗒 Shell

Claude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explaining complex code, and handling git workflow...

I am absolutely completely surprisingly shocked that the “age verification” data was being used to run dozens of background checks by a Thiel-related company. /s

www.malwarebytes.com/blog/news/20...

🔎 Claude Code Security, a new capability built into Claude Code on the web, is now available in a limited research preview. It scans codebases for security vulnerabilities. 💯
www.anthropic.com/news/claude-... #ClaudeCodeSecurity #ClaudeCode #Anthropic

hey people if you're ever looking to understand the architectual differences between nostr and atproto:
shreyanjain.net/2024/07/05/n...

open.substack.com/pub/leohohma...

techcrunch.com/2026/02/21/o...

The European Parliament pulls back AI from its own devices thenextweb.com/news/the-eur...

🎙️ Just published a new episode of Neural Newscast: Nvidia’s $30B OpenAI Stake and Trump’s 10% Global Import Tariff. Have a listen:

Welcome to life under the GOP. Still like their anti-woke policies or perhaps the liberal Democrats a don't seem so bad now?

I guess they don't believe in a smaller government.

DHS Wants a Single Search Engine to Flag Faces and Fingerprints Across Agencies | WIRED www.wired.com/story/dhs-wa...

www.el-balad.com/6856499

Fulu co-founder Kevin O’Reilly emphasizes the importance of user control in ensuring security. He hopes this campaign will highlight the compromises many users unknowingly accepted when installing such monitoring systems.

www.wsj.com/us-news/law/...

Putin tightens surveillance screws with new russian internet law

⚓️ Inter NYET 👿

This post appeared under this Techmeme headline:

techcrunch.com/2026/02/18/m...

This comes together to show that LLMs are getting decent at finding vulns, and that means vuln finding in semi-hard targets like Ghostscript is increasingly feasible. shostack.org/blog/vuln-fi...

🤖 3 AI motoru. Sıfır kesinti.

Gemini · Grok · Groq

Biri düşerse diğeri anında devralır.
Analizin asla yarım kalmaz.

👉 t.me/bullbearlabs_bot

#AI #yapayZeka #BullBearLabs

FEMA has released a new (2025), free online training: IS-2901: Introduction to Community Lifelines.

Enroll here https://training.fema.gov/is/courseoverview.aspx?code=IS-2901&lang=en

Learn more about the IMTA and become a member today at: https://www.ahimta.org/

With the EU's Cyber Resilience Act, #SoftwareTransparency isn't optional. It's a global mandate.

We're thrilled to announce #SBOM pioneer @allanfriedman.bsky.social is joining the Anchore board to help nav... https://anchore.com/blog/anchore-welcomes-sbom-pioneer-dr-allan-friedman-as-board-advisor/

macOS's Little-Known Command-Line Sandboxing Tool (2025)
https://igorstechnoclub.com/sandbox-exec/
[comments] [105 points]

Anthropic is launching Claude Code Security, a new cybersecurity capability for Claude Code, its popular AI coding tool.

RE: https://infosec.exchange/@BleepingComputer/116109509539088367

@iFixit and here I was under the impression that the indicator lights are hardwired (no software/firmware) - as soon as the mic or camera receive power, the lights turn on.

One provision would have required businesses to bar people from accessing their sites via "a virtual private network system or virtual private network provider." www.cnet.com/tech/service...

LOL nope.

www.macrumors.com/2026/02/20/j...

People need to pay attention to the differences in framing between this WSJ article and the CTV article by @andrewweichel.bsky.social. The WSJ clearly states that OpenAI staff raised the chats to management and they chose not to alert RCMP. The CTV article does not. www.ctvnews.ca/vancouver/ar...

How digitally sovereign are you? Red Hat can help measure that

Red Hat's toolkit offers governments and enterprises a way to measure the control they actually have over their data, infrastructure, and operations in this era of geopolitical cloud anxiety.
#hackernews #news

YOU CAN'T BE TOO CAREFUL #Hackers #DarkWeb #CreditCards

I quit ChatGPT — here's how I moved everything to Claude and Gemini without losing my data (or my mind)

Read more on: tom-s-guide.visitlink.me/uplWPi

Selling AI through the cloud comes at a price: Anthropic could share roughly a tenth of projected revenue with partners.

Nitrogen Ransomware Bug Locks Out Attackers from Victims' Data #Coveware #ESXiStrain #NitrogenRansomware

I just tried Grok, it doesn't mind any of these words unlike Copilot and Gemini.

open.substack.com/pub/leohohma...

📌

Shadow Campaigns Expose 37 Nations to State-Linked Cyber Espionage Operations #CriticalInfrastructure #criticalinfrastructurecybersecurity

🎮 **Hackers Raise The Alarm About Discord’s Recent Age-Verification Partner**

Discord’s contested global age-verification roll-out keeps sounding worse

📰 Source: Kotaku
🔗 Link: https://kotaku.com/discord-age-verification-hack-persona-2000672570

#AI #ArtificialIntelligence

11 Useful Raspberry Pi Projects With No Extra Hardware
#raspberrypi #linux #opensource
raspberrytips.com/raspberry-pi...

canartuc.medium.com/10-billion-d...
Great story about an unsung software developer and a great tool he maintains.
I use curl practically every day and always took it for granted before the story

NVIDIA releases open-source robot world model trained on 44,000 hours of human video

www.perplexity.ai/page/nvidi...

@NVIDIA #AI #MWC26

AI vulnerability to cyberattacks, and manipulation.

Operation DoppelBrand Weaponizes Trusted Brands For Credential Theft <- phishing lures redirec victims to counterfeit pages where credentials are harvested and transmitted to Telegram bots

I love initiatives like Claude Code Security (even if it's gated for now). But I'm also terrified of them.

claude.com/solutions/cl...

" .. Anthropic is refusing to let the government use the company’s technology for autonomous weapons & domestic surveillance. In response, the Pentagon is threatening to cut business ties and declare Anthropic a 'supply chain risk.' Who will blink first?" www.nytimes.com/2026/02/20/p...

I had the horrifying realization that some candidates' written answers to standard questions from reporters were written by AI.
Still worth reading.
#ncpol

@realtexaspaul.com 👇🏿 👀
🤙🏻🖖🏻
#OpenAI is Suddenly in #Trouble
95,247 views · 2 hours
youtu.be/-q2n5DkDoMQ

Android Central より

Google just doubled its AI reasoning power with the surprise launch of Gemini 3.1 Pro
GoogleはGemini 3.1 Proの驚くべき発売でAIの推論力を倍増しました

Android Central より

Should you buy the Google Pixel 10a? Well, the answer isn't that simple
あなたはGoogle Pixel 10aを購入するべきですか? まあ、答えはそれほど簡単ではありません

They should've gone for an Apple II-based game board like Tic-Tac-Dough. www.museumofplay.org/blog/apple-i...

Reddit, Meta, and Google Voluntarily Gave DHS Info of Anti-ICE Users, Report Says

gizmodo.com/reddit-meta-...

Greenleaf AI Studios Membership has its Privileges #ai #model #girl #video #shorts

'Some employees interpreted Van Rootselaar’s writings as an indication of potential real-world violence, and urged leaders to alert Canadian law enforcement about her behavior, the people familiar with the matter said.
OpenAI leaders ultimately decided not to contact authorities....'

DHS and ICE are using AI-powered surveillance tools like Mobile Fortify beyond the border, integrating facial recognition and biometric monitoring into interior enforcement operations. #Trump #ImmigrationSurveillance #DHS #ICE #MarshallProject

Apple is often leading the way

9to5mac.com/2026/02/21/h...

Anthropic rolls out embedded security scanning for Claude | CyberScoop

Big Brother wants to cage you like an animal. They need tech to do it faster

www.wired.com/story/dhs-wa...

"Identity verification efforts like these mislead people into thinking you’re more protected. Computer scientists have long sounded the alarm that a central database of thousands of ID documents will always be a lucrative target for attackers."

This is a big deal for basically any/all manufactured tech.

#hdd #costofai #economy

www.androidheadlines.com/2026/02/west...

its so funny how other companies drama is like "we killed another studio because we hate them, also you better believe AI is the future" meanwhile the apex of Nintendo drama is

www.malwarebytes.com/blog/news/20... @dunkindonutscoffee.bsky.social @icij.org @carnivalcruise.bsky.social @deptofinterior.bsky.social @amyklobuchar.com @oigdc.bsky.social @msnowreports.bsky.social @aliciamenendez.bsky.social

Burn AI to the ground.

Experts recommend placing a sticker on your credit card to obscure sensitive info like the CVV and expiration date, deterring visual theft and fraud. #CreditCard #News

Why must healthcare embrace Agentic AI for data protection

Are Non-Human Identities the Key to Unlocking Agentic AI in Data Protection? Organizations across industries are increasingly focusing on the management of Non-Human Identities (NHIs). These machine identities, akin to di…
#hackernews #news

RE: https://infosec.exchange/@briankrebs/116108996488864099

I trust anything funded by Thiel not a whiff, not a mile, not at all.

cybernews.com/security/mic...

Exclusive: OpenAI just raised its revenue outlook—but now expects to burn $111B more cash by 2030. The AI boom’s economics are getting clearer.

No wonder so many people hate AI. They can snitch on crimes!

www.theguardian.com/world/2026/f...

App X tracks users even with data‑sharing off—collecting device IDs, IPs & browsing behavior. Limit exposure: blockers (uBlock/Privacy Badger), Incognito, disable JS, turn off personalized ads. #Privacy #Surveillance #DataTracking #DigitalRights

ℹ️ tinyurl.com/mpdheawt

📰 Google's Gemini 3.1 Pro Tops AI Benchmarks at One-Third the Cost of Competitors

Google has unveiled Gemini 3.1 Pro, a new AI model that leads the Artificial Analysis Intelligence Index while reducing inference costs by up to 67% compared to rival models. Experts cautio...

#AINews #AI #Teknoloji

Honestly no wonder all of California is full of new and emerging defense tech startups

Gemini 3.1 Pro Unveiled: AI Crushes Benchmarks!
#Gemini3Pro #AIBenchmarks #FutureTech
www.squaredtech.co/gemini-3-1-p...

PromptSpy is the first known Android malware to use generative AI at runtime

Researchers have discovered the first known Android malware to use generative AI in its execution flow, using Google's Gemini model to adapt its persistence across different devices. [...]
#gemini #hackernews #news

Saw some ask Google Gemini AI to do a 3d render of an ocean as if the Ocean Modifier isn’t a default tool in Blender

If you feel like you don't adequately understand just how problematic surveillance tech and """age verification""" laws are, let this be an indicator.

(It's not about verification, it's about identification, which comes with drastic personal privacy/security costs.)

www.therage.co/persona-age-...

www.theglobeandmail.com/canada/artic...

An extension of Anthropic's contract with the Pentagon is being held up over talks on additional guardrails the AI company wants to put on Claude Gov, which it built for the U.S. national security apparatus.
www.stripes.com/theaters/us/...

📰 Gemini 3.1 Pro Preview Tops AI Leaderboard with Unprecedented Cost Efficiency

Google's Gemini 3.1 Pro Preview has emerged as the new leader in the Artificial Analysis Intelligence Index, outperforming rivals like Claude Opus 4.6 by 4 points while costing less than half...

#AINews #AI #Teknoloji

If you harbor any skepticism whatsoever about this technology and the companies profiting off it you're a smug luddite who has no idea how beneficial it's going to be

Benedict Evans on the difficulties OpenAI and other foundational model creators face when having to generate revenue:

www.ben-evans.com/benedictevan...

Great essay.

militarnyi.com/en/news/ukra...

Google's Gemini 3.1 Pro Doubles Reasoning Performance and Retakes the AI Crown



#Google #Gemini #Benchmarks

A new bill could finally bring parity for WATEC's cybersecurity employees, but concerns over hiring flexibility loom large.

Learn more here

#WA #CitizenPortal #TechEquity #LaborRights #PublicPolicy

Citizen Lab: Kenyan authorities used Cellebrite to break into phone of dissident therecord.media/spyware-keny...

Measuring AI agent autonomy in practice \ Anthropic

This Techie explains why/how AI Agents represent security threats: allowing Agents to purchase for you; help with your emails; etc. - it exposes you to serious breaches.

youtu.be/TdHg9ee56Iw?...

Interested in improving the appearance and usability of your open-source repo? Nikolay Nikitin presents an agentic-AI-powered tool he's developed (also open source!) for precisely this task.

According to Ransomware.live, thegentlemen ransomware group has added Grupo Progresso (🇧🇷) to its victims.

According to Ransomware.live, thegentlemen ransomware group has added Seac (🇮🇹) to its victims.

"As far as I can tell the proliferation of the technology mainly benefits the people making money off of it, whereas, say, a robust and fast train network would provide a lot more obvious benefits to working people in the country where I live."
#AI #Anthropic
aftermath.site/anthropic-cl...

#law #lawsuits #AI #MentalHealth

'A Georgia college student named Darian DeCruise has sued OpenAI, alleging that a recently deprecated version of ChatGPT “convinced him that he was an oracle” and “pushed him into psychosis.”'

arstechnica.com/tech-policy/...

According to Ransomware.live, thegentlemen ransomware group has added Line Up Korea (🇰🇷) to its victims.

According to Ransomware.live, thegentlemen ransomware group has added Sasin Colleges & Universities (🇹🇭) to its victims.

Crims create fake remote management vendor that actually sells a RAT

$300 a month buys you a backdoor that looks like legit software
Researchers at Proofpoint late last month uncovered what they describe as a "weird twist" on the growing trend of criminals abusing remote monitor…
#hackernews #news

Write to Keir Starmer to demand that he to cancels all government contracts with Palantir, which provides AI technology used in Israel's genocide in Gaza palestinecampaign.eaction.org.uk/cancelpalant...

Why using facial recognition on your phone could leave you vulnerable

Journalists and activists aren’t the only ones who can have their #Biometrics used against them. Here’s how to protect yourself.

www.fastcompany.com/91486312/why...

#CyberSecurity

Gemini 2 Flash Thinking vs Grok 4.1 — totally different approaches to prompting chains.

Lattice-Based Zero Trust Identity Verification for AI Agents

Explore lattice-based zero trust identity verification for AI agents. Secure MCP deployments with quantum-resistant encryption and 4D access control.
#hackernews #news

Spyware can sneak into your smartphone and track your activities -- checking for unusual activity and unfamiliar apps can help protect your device.

OpenAI is Suddenly in Trouble youtu.be/-q2n5DkDoMQ?... via @YouTube

Listen I know people hate ai and it's evil but my PC wouldn't boot up and I just took a picture of my PC, typed into Google Gemini and said, this PC won't boot what do I do?

And it told me how to fix it

Supply Chain Attack Secretly Installs OpenClaw for Cline Users www.darkreading.com/application-...

the folks at cloudflare often come up with elegant solutions (although in this case it's mostly a smart implementation of an idea floated by anthropic a couples of months ago)

blog.cloudflare.com/code-mode-mcp/

meanwhile..😲DHS stopped funding last year to agencies responsible for creating security plans for events w/ large crowds against terrorist threats, etc-bc it No Longer Served the Mission of DHS.-Wasn't that what DHS was created for in the 1st place⁉

www.nytimes.com/2026/02/20/s... 🎁

📰 Google's Gemini 3.1 Pro Stumbles on Technical Image Generation Test

A new benchmark test reveals Google's latest Gemini 3.1 Pro AI model continues to struggle with generating technically accurate images, despite claims of leadership in the field. The test, focused on a...

#AINews #AI #Teknoloji

Gen AI would be a beautiful name for a baby android

Electro Optic Systems opened a Singapore facility to meet the growing demand for high-power laser weapons and defense systems.

The company says the new development represents a milestone in its global growth strategy.

Read the OPN story: https://bit.ly/3MBc500

#lasers #defense #technology 💡 ⚛️

📰 Beyond the Browser: Power Users Reveal Advanced AI Workflows

A Reddit user's plea for help reveals a growing divide between casual and power users of AI tools. While official support channels like Google's YouTube Help offer basic troubleshooting, a community of advanc...

#AINews #AI #Teknoloji

Russia stepping up hybrid attacks, preparing for long standoff with West, Dutch intelligence warns therecord.media/russia-cyber...

Quote of the day by Alan Turing: ‘Be kind, have a sense of humour, fall in love, enjoy strawberries’ - The Economic Times share.google/yJYl6tGA15IC...

keepandroidopen.org

The going ones in the US, UK, China, and India are a dire warning to the world about the dangers of big tech embedded surveillance into our daily lives. Strong safeguards must be put in place to stop the Orwellian neo-Stasi states from establishing absolute control over our lives

Ugh, it must be my anti AI posts, Google smelling what I'm typing, because I just sent an email submission to a journal WITHOUT A SUBJECT LINE, something Gmail has not previously permitted me to do. LOL

Please note this automatically generated data as evidence of spyware being used on the phone to make fake narratives. You can match the pattern from yesterday and today I’m plugged into the same power plug to see today. Spyware is being used to slow charging on the phone @ Apple

www.forbes.com/sites/daveyw...

www.theglobeandmail.com/canada/artic...

www.bleepingcomputer.com/news/securit...

Runtime Tracing for AI Agents: What Your OpenClaw Agent Actually Does Inside the Container by Mario Candela itnext.io/runtime-trac...

SBOM-first isn't just a buzzword—it's the architecture that makes continuous security actually possible 🔄

Feel the difference ⚡

https://anchore.com/platform/

#SBOM #CRA #SoftwareSupplyChain #Compliance

Google says its AI systems helped deter Play Store malware in 2025 techcrunch.com/2026/02/19/g...

Obedient Traders Respond to Claude Code Cybersecurity Plugin by Selling Cybersecurity Stocks #Technology #Cybersecurity Other