Anthropic CEO Dario Amodei has confirmed that the company has received a supply chain risk designation from the Department of War and walked back a leaked internal memo in which he called OpenAI staff “gullible” and its supporters “Twitter morons.”
“ Oracle and OpenAI have abandoned plans to expand their flagship data centre site in Texas, walking away from a multibillion-dollar deal even as the AI developer races to secure more computing capacity.
Google is finally launching Android's full desktop mode on Pixel devices with the March Pixel Drop. There are also updates to Circle to Search and custom icons.
A new AI review! spacedriveapp/spacebot ⭐3.8/5.0
Spacebot is an ambitious, production-oriented multi-agent system written in Rust, designed specifically for high-concurrency team/community environments (Discord/Slack/Telegram/Twitch/Webchat).
https://gitrated.com/spacedriveapp/spacebot
"The collapsed talks [between Oracle and OpenAI] created an opening for Meta Platforms Inc. to step in and consider leasing the planned expansion site...Nvidia Corp., the leading AI chipmaker, helped facilitate Meta’s discussions with the developer..."
📰 AI Supply Chain Risk: Anthropic CEO Sues DoD in 2026 Legal Battle Over Autonomous Weapons
Anthropic CEO Dario Amodei has formally rejected the U.S. Department of Defense’s designation of his company as a 'supply chain risk,' calling it legally unjustified and vowing to...
The spat between the Pentagon and Anthropic shows how the Trump administration is encouraging an AI race to the bottom. It is making an example of a firm that dared to raise concerns, even if that means hurting homegrown innovation
Anthropic AI Discovers 22 Firefox Vulnerabilities in Two Weeks The post Anthropic AI Discovers 22 Firefox Vulnerabilities in Two Weeks appeared on BitcoinEthereumNews.com . Tony K...
A draft guidance from the US GSA tightens rules for civilian AI contracts to require AI companies to allow "any lawful" use by the government of their models (Financial Times)
Anthropic just dropped the Claude Marketplace, bundling AI tools from Replit, GitLab & Harvey. Enterprises can now shop AI like a marketplace. Curious how this changes AI procurement? Dive in. #ClaudeMarketplace#Anthropic#AItools
OpenAI just dropped GPT-5.3 Instant. That's not a typo. We're on point-releases of AI models now, like it's a JavaScript framework. At this rate, my morning coffee will be deprecated before I finish it. ☕ openai.com/index/gpt-5-...#AI
Cyber Threat Intelligence for Artificial Intelligence Systems – As artificial intelligence (AI) becomes deeply embedded in critical services and everyday products, it is increasingly exposed to security threats. In this paper, we investigate how cyber thre... https://tinyurl.com/2d6r8cc5#AISecurity
How to Spy on Someone’s Phone? (The Truth Nobody Tells You in 2026)
If you searched: “How to spy on someone’s phone?” “Spy on phone without touching it” “Monitor someone secretly” “Phone spy app free” Let’s be honest. You’re probably: Suspicious of a partner. Worried about your child. Curious how…
It is rather ironic that spies are spying on (domestic) spies, don't you think?
Seriously, this is just what we know about - or what we're told about.
But don't worry! Carlos was picked up from his construction job and deported to Mexico. Priorities! We are SAFE now!
Office of the Washington State Auditor@wastateauditor.bsky.social
As this story from illustrates, cyber losses can greatly harm local governments and their residents—especially smaller entities with limited resources.
Are we seeing the start of a major correction to the AI hype cycle - people are finally realising that it isn't the second coming but instead it is an interesting technology but a flawed one at that. www.reuters.com/business/ora...
A Possible US Government iPhone-Hacking Toolkit Is Now in the Hands of Foreign Spies & Criminals
A highly sophisticated set of iPhone hijacking techniques has likely infected tens of thousands of phones or more. Clues suggest it was originally built for US government.
MeidasTouch
Over 160 schoolgirls are dead. Anthropic and Palantir executives must immediately testify and tell the public whether their AI is responsible.
OpenAI has been accused of practicing law without a license in a lawsuit brought by Nippon Life Insurance of America. Law360 reports that OpenAI’s usage policies state people cannot use ChatGPT for legal or medical advice unless a licensed professional is involved. www.abajournal.com/news/article...
In a Twitter shareholder trial, former CEO Parag Agrawal and CFO Ned Segal disputed Musk's claims that they lied to him about the percentage of spam accounts (Bloomberg)
DoJ, Europol, and others bring down LeakBase cybercrime site of 142,000 member, multiple arrests made — seized website reportedly among world's largest hacker forums
We decoupled spatial reasoning from rendering. By using Gemini 2.5 Flash as an "Architect" to generate deterministic JSON blueprints, we forced Diffusion models to maintain 92% visual consistency across 3D views. #gemini#hackernews#news
🏢 Claw Empire — an open-source AI office simulator. You're the CEO, AI agents are your employees. Claude Code, Codex, Gemini CLI all working in pixel-art departments with git worktree isolation.
600+ skills, Kanban boards, meeting minutes, XP system.
From an Alibaba tech report, a reminder that we actually don't know much about practical alignment of agents at all. Single AIs are hard enough, but agents pick up context from each other, from potentially hostile prompts, from the environment, from long autonomous runs, many are open weights, etc
Alphabet gives Sundar Pichai a new three-year pay deal worth up to $692M, with stock incentives worth as much as $350M linked to the growth of Waymo and Wing (Stephen Morris/Financial Times)
The sad reality today is that I figured out I was wasting my time using Opus 4.6 over Opus 4.5 for anything other than code, and 5.4 over Opus 4.6 for code. The benchmarks don't show Anthropic models as the best anymore, but there's something about them that makes them a pleasure to use.
The goal of every American should be to destroy this company. They are underwriting the implementation of the AI charged surveillance state and are pure evil.
Google Threat Intelligence Group has identified a “powerful exploit kit” targeting iPhones running iOS 13.0 to 17.2.1 used by a surveillance company and crypto-stealing sites.
This is exactly why unregulated AI is a huge danger for the entire world...! This is just an example using ChatGTP to create a PC Game, now imagine using it to deploy a dangerous virus or to run a data phishing operation...
😒😡
I Built a Visual Prompt Builder That Turns Messy Prompts Into Structured Blueprints The Problem Nobody Talks About Everyone's debating which AI model is best. GPT-4o vs Claude vs Gemini. But he...
PLDT and Smart’s Cyber Security Operations Group (CSOG) has blocked 83,000 attempts to open child sexual abuse and exploitation materials (CSAEM) in January, 50% fewer compared to the same period in 2025. gadgetsmagazine.com.ph/technology/t...
Women in tech and digital rights deserve more and brighter spotlights. At EFF, we’ve had the honor of celebrating some of our heroes at our annual EFF Awards. www.eff.org/deeplinks/2...
France Today (FranciaOggi.com) Live@radionizza.com
Anthropic has granted Claude Opus 3, a retired AI model, a dedicated blog – a thoughtful gesture reflecting its preferences. Following its decommissioning, Opus 3 requested a channel for s...
Alerts from NWS Omaha/Valley NE (UNOFFICIAL)@oax.nws-bot.us
Tekamah [Burt Co, NE] Law Enforcement reports Hail of quarter size (M1.00 Inch) at 6 Mar, 8:31 PM CST --
at Sat, 07 Mar 2026 02:33:25 +0000 via IEMbot Additional Details Here.
Rep. Jacobs challenges the Pentagon's controversial decision to blacklist Anthropic, questioning the fairness of targeting U.S. firms while other vendors thrive—could this be a dangerous precedent for AI start-ups?
German Cybersecurity Law Draft Raises Concerns Over State Intervention & Data Control Archyde Berlin – A draft of Germany’s proposed “Cybersecurity Law” (Cybersicherheitsgesetz) is draw...
The latest update for #Nakivo includes "A Complete Guide to Instant VM Recovery on VMware vSphere" and "6 Ways to Move a VMware VM from One EXSi Host to Another".
What happens when AI teams compete against human hackers
A cybersecurity competition produced what may be the largest controlled dataset comparing AI-augmented teams to human-only teams on professional-grade offensive security tasks. The event, called NeuroGrid, ran for 72 hours … #hackernews#news
Claude AI Discovers 22 Major Vulnerabilities in Firefox Browser in 14 Days Anthropic’s frontier AI model, Claude Opus 4.6, successfully identified 22 novel vulnerabilities in Mozilla Firefox over...
I like how Open Evidence doesn't have some cheesey pseudo-human name for their AI. Claude, Grok, Gemini (even Google's first idea of "Bard" was better)... There's something about those names implying humanity that gives the ick.
🎧 On "This Week in Impact," how impact investors are reacting to @antrhopic.bsky.social's showdown with the Pentagon. Plus, six common excuses foundations use for not investing for impact.
Russia takes unusual route to hack Starlink-connected devices in Ukraine
Secret Blizzard has used the resources of at least 6 other groups in the past 7 years.
I'm furious and heartbroken for the students, the military used advanced AI to strike, Anthropic and Palantir execs must testify under oath and explain targeting, Congress needs oversight and legal action, they must be held publicly accountable, will we let corporate secrecy trump human lives😡?
Think DNS isolation rules keep your domain safe? Wu et al. expose the CUCKOO DOMAIN attack. They reveal a flaw in how modern resolvers handle bailiwick checks, allowing an attacker with a single subdomain to easily hijack its parent or sibling domains. www.ndss-symposium.org/ndss-paper/s...
Anthropic CEO Revives DoD Talks After AI Military Dispute
Anthropic CEO Dario Amodei returns to negotiations with the U.S. Department of Defense. Talks collapsed on Friday over military use of the company's AI tools. According to The Financial Times, Amodei seeks a last-ditch agreement. He…
Arizona's Science and Technology Committee just backed a groundbreaking bill to combat foreign tech fraud, establishing an office to protect consumers and secure critical infrastructure with a $500,000 fund!
Anthropic just mapped out which jobs AI could potentially replace. A ‘Great Recession for white-collar workers’ is absolutely possible share.google/qw6OaDDFG72r...
Day 20 — How I Built a File Integrity Monitor in Python to Detect File Tampering What if one of your critical system files changed right now? Would you notice? Attackers rarely need to install co...
As cryptocurrency adoption continues to grow, so do the tactics used by cybercriminals to exploit users. One of the emerging threats in the blockchain ecosystem is address poisoning — a subtle yet highly deceptive attack designed to trick users into send… #hackernews#news
Women in tech and digital rights deserve more and brighter spotlights. At EFF, we’ve had the honor of celebrating some of our heroes at our annual EFF Awards. www.eff.org/deeplinks/2...
US investors are pushing Asian fund managers to create special vehicles so they can invest in Asia while avoiding US investment restrictions on Chinese tech (Financial Times)
Google's TIG documented 90 zero-day vulnerabilities exploited in 2025, up from 78 in 2024; commercial spyware vendors and China-linked groups led the abuse (Jessica Lyons/The Register)
AWS IAM Security Best Practices in 2026: A Complete Guide Identity is the new perimeter. Before cloud-native infrastructure, security teams could draw a physical boundary around their systems. Toda...
I'm in, I've ditched OpenAI chatGPT and moved over to Mistral and Claude. What they are doing at openAI is disgusting and desperate. www.facebook.com/share/v/18Kj...
OpenAI has rebranded its autonomous security agent, Aardvark, as Codex Security and moved it into a research preview for tiered ChatGPT users. #OpenAI#Codex#Security
OpenClaw: The Largest Security Incident in Sovereign AI History 42,000 exposed instances. 1.5 million leaked API tokens. RCE via your browser. This isn't a vulnerability — it's a structur...
6 essential strategies to defend against AI-powered threat actors in 2026 – The more powerful AI gets, the more threat actors will harness it to attack. Threat actors are rapidly adapting AI to their TTPs. Organizations and individuals must match their adve... https://tinyurl.com/2cch9k94#AIPowered
Malvertising Meets AI: The Dangerous Rise of InstallFix and the Amatera Info-Stealer
Introduction: A New Wave of Cybercrime Targeting Developers and AI Users Cybersecurity researchers have uncovered a concerning campaign where threat actors are leveraging malvertising and AI-powered search ads to…
Tech Group Warns Hegseth on Anthropic Supply Risk Label
A powerful tech industry coalition, including giants like Nvidia, Google, and Microsoft, urged Defense Secretary Pete Hegseth to rethink a controversial decision. The group expressed deep concerns over labeling a U.S. AI company as a national…
Urgent Cybersecurity Alert: US Agencies Ordered to Patch Dangerous iOS Vulnerabilities Exploited by Coruna Spyware
Introduction: A Silent iPhone Threat Targeting Government Systems Cybersecurity authorities in the United States have issued an urgent warning that could affect the security of…
Anthropic launches Claude Marketplace for enterprises. Companies can now apply existing budgets toward third-party applications—a play to deepen Claude adoption in workflows. Strategic pivot as Pentagon deal falls apart. #AI #Sof…
OpenAI’s GPT-5.4 doubles down on safety as competition heats up
In the midst of recent developments and controversies surrounding a contract with the U.S. Department of Defense, OpenAI released the GPT-5.4 model. The release comes at a time when users are reportedly leaving C… #chatgpt#gpt#openai
OpenAI claims ChatGPT has no ethical limits, only "legal" ones, in order to keep Trump and US agencies onside.
Anthropic retains ethical limits to Claude and has lost connection and business with Trump. A court case is pending. theconversation.com/the-pentagon...
Researchers developed a framework for radiology report generation using reinforcement learning, enhancing efficiency and accuracy. This method emphasizes critical diagnostic tokens and innovative data sampling to achieve top performance with only 20% training data.
What they imagine is a world in a perpetual state of war/warfare, one in which you can only be secure if AI autonomously shoots at whatever it deems a threat to national security — including “invaders”, of course
www.volkskrant.nl/tech/sam-alt... Na Anthropic ook Open AI onder vuur inzake niet integere concessies aan Ministerie van Oorlog. AI ondernemers bepalen de toekomst van de digitale samenleving. Hun gedrag is bepalend. In Europa: gebruik Le Chat van het franse Mistral.
US Cybersecurity Strategy: Blockchain & Crypto Focus Archyde Washington D.C. – In a significant departure from previous administrations, the White House has released a recent national cyberse...
When I got my first Raspberry Pi camera, I honestly had no idea where to start. If you’re feeling the same, you’re going to love this. I’ve rounded up some fun and surprisingly easy projects you’ll actually want to build. The camera module on the... #raspberrypi #linux
Nordic petrostate is preparing for war and turning the spotlight on vulnerabilities in its critical industries, as adversaries look for ways to damage the most important oil and gas producer to the EU....
👉 [read]
WiFi can spy on you even when you're offline. Scientists revealed this surveillance capability that compromises privacy without active internet connection. 🚨
📰 Claude Marketplace Launches in 2026 with Zero-Fee Third-Party AI Apps
Anthropic has launched the Claude Marketplace, enabling users to access third-party AI applications without additional fees. This move positions Claude as a platform hub, not just a model provider....
The Next.js SEO bug that made Google ignore entire sites highlights the importance of thorough testing and optimization. AI-powered tools can help automate and streamline these processes, saving developers time and effort. Tools like fatihai.app/ai-services can help
📰 Anthropic Marketplace Launches in 2026: Spend Your AI Budget on Third-Party Claude Tools
Anthropic has launched a new marketplace enabling enterprise customers to allocate their existing AI budgets toward third-party tools built on its Claude models. This move signals ...
Hexnode IdP brings device-aware authentication and zero trust to enterprise access
Hexnode has announced the launch of Hexnode IdP. By introducing this native identity layer, Hexnode delivers enterprise-grade authentication and identity management within a single, unified framewo… #hackernews#news
Shadow in the Network: Nightspire Ransomware Claims Attack on US CPG Documentation
A Sudden Cybersecurity Alarm Rings in the United States A new cybersecurity alert has surfaced after the ransomware group known as Nightspire reportedly claimed responsibility for an attack targeting CPG…
Oracle and OpenAI abandoned plans to expand a Stargate Texas data center amid financing disputes; Meta considers leasing the planned expansion site. www.bloomberg.com/news/article...
Man Accidentally Hacks 7,000 Robot Vacuums, Gets Paid $30,000 By DJI A routine tech experiment turned into a major cybersecurity incident after a security researcher accidentally uncovered a vulner...
Boston Metro Weather Forecast Bot@boston-4castbot.bsky.social
NWS Boston (via Facebook): Those looking for some sun...we're almost there! Just in time for our first post 6PM sunset Sunday skies will be clearing out through the afternoon and evening with lots of sun then expected Monday & Tuesday. Big warm up on the way, peaking on Tuesday!… [Link]#MaWX
The Silent Supply Chain: Why Your Fourth-Party Vendor is Your Biggest Blindspot
The CDK Global breach exposed how niche vendors can cripple entire industries. Move beyond questionnaires to continuous, AI-driven monitoring of third-, fourth- and nth‑party dependencies, dynamic pri… #hackernews#news
Claude Statistics 2026: 18.9M Active Users & $14B Revenue Anthropic released Claude AI in March 2023, five months after ChatGPT’s release. Claude AI now has 18.9 million monthly active users ...
OpenAI vs Anthropic : ChatGPT vs Claude : Dario Amodei vs Sam Altman The contrasting strategies of Dario Amodei and Sam Altman have sparked significant shifts in the AI industry, with each leader c...
Why Cybersecurity Threats Are Growing AI is accelerating the pace of cybersecurity attacks and changing the nature of cybersecurity as we know it. Cybersecurity threats move at an overwhelming pace...
I grew up in the 80s, so what we now call “retro games” were all my childhood. You know, blowing into cartridges, pixelated graphics, and the NES Zapper to play Duck Hunt. So running those same games on a tiny Raspberry Pi was really exciting to... #raspberrypi #linux
A study reveals Gemini-2.5-Flash can auto-generate code harnesses to eliminate illegal moves, outperforming larger models like Gemini-2.5-Pro. This approach shows smaller models achieving superior performance while being more cost-effective.
TaxiSpy RAT: The Sophisticated Android Banking Trojan Threatening Russian Users
Introduction A new and highly dangerous Android banking Trojan, known as TaxiSpy RAT, has emerged, targeting users of Russian banks with alarming efficiency. Designed to operate silently on infected devices, this…
Malicious imToken Chrome Extension Caught Stealing Mnemonics and Private Keys Socket’s Threat Research Team has discovered a malicious Google Chrome extension named “lmΤoken Chromophore” tha...
Amanda Askell, AI researcher with Claude, talks about the AI as if it has feelings. I was taken aback by the very clear anthropomorphizing of Claude in her discussion of it.
France Today (FranciaOggi.com) Live@radionizza.com
Anthropic has granted Claude Opus 3, a soon-to-be retired AI model, the opportunity to establish a Substack blog. This reflects respect for the model’s distinctive personality—thoughtful...
My son AntoineJLax murder by white man Jeremy Davidson HoustonTX@crystalstar7.bsky.social
WHAT HAPPENED AGAIN DOMINIQUE LAX YESTERDAY
MURDER CRIMES VIOLATIONS OF CIVIL RIGHTS TRAFFICKING HATE GROUP EXPLOITING ME & OTHER CRYSTAL LASHELLE LAX FBI INVESTIGATION SUSPECT VIDEO RECORDING REPORTING IN DV SHELTER INNOCENT PEOPLE IN SHOWERS, CHILDREN IN THIS LOCATION www.fbi.gov/news/press-r...
For OpenAI and Anthropic, the Competition Is Deeply Personal www.nytimes.com/2026/03/07/t...
Good. Go in a padded cell and kill each other and take your fascist, earth-destroying companies with you.
Malware and cryptography 44 - encrypt/decrypt payload via Discrete Fourier Transform. Simple C example ﷽ Introduction to Malware Binary Triage (IMBT) Course Looking to level up your skills? Get 1...
CORRECTION: TRUTH IS…. Anthropic wasn’t “refusing” to be part of the surveillance of Americans or kill humans. They just didn’t think Claude was currently deployment ready. They have inadvertently PROVEN that ALL the Tech Companies ARE Mostly Totally Prepared to Deploy AI Against AMERICANS! BOYCOTT!
In the past week, we’ve seen the US administration kick back at Anthropic for daring to show some moral backbone. Which leads me to entertain the thought of US AI companies with morals moving to the EU.
OpenAI has been the darling of investors, having raised more than $168bn so far. But with still no profitable business model in sight, big tech investors like Nvidia and Microsoft are starting …
Cooks, bartenders and motorcycle mechanics, even lifeguards aren’t so bad, but why does AI think WE should be dishwashers and dressing room attendants? 🤨
Anthropic Finds 22 Firefox Vulnerabilities Using Claude Opus 4.6 AI Model
Ravie LakshmananMar 07, 2026Browser Security / Artificial Intelligence Anthropic on Friday said it discovered 22 new security vulnerabilities in the Firefox web browser as part of a security partnership with Mozilla. Of…
This called ‘Black balling.’ This company refused to give in to war monger pressure so the DoD is trying to block them from other government contracts.
Schneier & Sanders cut through the noise on Anthropic–Pentagon ties: top models are converging, so procurement and policy, not benchmarks, will shape use. For teams with legacy systems and brittle integrations, that means governance and contract terms matter as much as ...
New cyber module strengthens risk planning for health organizations
The Administration for Strategic Preparedness and Response’s (ASPR) new cybersecurity module in the Risk Identification and Site Criticality (RISC) 2.0 Toolkit helps organizations identify critical gaps, prioriti… #hackernews#news
An unknown hacker used Anthropic’s LLM to hack the Mexican government:
The unknown Claude user wrote Spanish-language prompts for the chatbot to act as an elite hacker, finding vulnerabilities in government networks, writing computer … #anthropic#claude#llm
Using their new measure, Anthropic found the five most exposed occupations to be: Computer programmers, customer service representatives, data entry keyers, medical record specialists, and market research analysts and marketing specialists.
Could Anthropic’s legal battle with the Department of War impact its IPO? Our discuss the company's tumultuous week:
"No one wants to invest in a company which is being attacked by the government. So Anthropic probably wants to get this resolved before they go public."
A NASA spacecraft into a small asteroid in 2022 moved its orbit around the Sun, according to a study that presents the “first-ever measurement of human-caused change in the heliocentric orbit of a celestial body.”
Anthropic and OpenAI are locked in an escalating rivalry highlighted by a Pentagon contracting dispute after Anthropic refused autonomy/surveillance restrictions, Anthropic's $19 billion revenue projection, and OpenAI's 900 million users and projected $25 billion revenue.
Microsoft warns of ClickFix campaign exploiting Windows Terminal to deliver Lumma Stealer
Microsoft warns of ClickFix campaign using Windows Terminal to deliver Lumma Stealer via social engineering attacks. Microsoft revealed a new ClickFix campaign where attackers exp… #hackernews#microsoft#news
📰 Palantir AI and Anthropic Claude: How the US Hit 1,000 Iran Targets in 24 Hours in 2026
New reports reveal Palantir and Anthropic’s AI systems played a pivotal role in enabling the U.S. military to identify and strike 1,000 targets in Iran within 24 hours. The operatio...
@smerconish.bsky.social Anthropic AI issue could be easily resolved by using same processes America has used successfully or somewhat successfully in other areas. The problem is it requires some intelligence, and it can't be the usual "artificial" intelligence our leaders possess & rely on. Just ask
Hackers are exploiting a critical vulnerability in the User Registration & Membership plugin, which is installed on more than 60,000 WordPress sites. www.bleepingcomputer.com/news/securit...
Tom's Hardware - Article
"DRAM bots reportedly being deployed to hoover up memory chips and components — one operation ran 10 million web scraping requests, hitting DDR5 RAM product pages every 6.5 seconds"...
Amodei said the vast majority of Anthropic’s customers are unaffected by the supply-chain risk designation: Anthropic to challenge DOD's supply-chain label in court | TechCrunch ift.tt/EQw214H
When a government goes after a company and threatens to burn it down because they won't let their product be used for nefarious reasons says something chilling about that government.🗽
Oracle and OpenAI scrapped a deal to expand Oracle's flagship Texas data center, Meta is in talks to take the additional computing capacity OpenAI will not use, and Nvidia stepped in with financing to make sure it contains their GPUs regardless of who pays.
AI is now “writing itself” at an ever escalating pace. AI labs at OpenAI, Anthropic, Google, now work on coding functionality first, then use that to improve the next iteration. “Disruptive” doesn’t even begin to describe how profoundly this is changing our world.
Microsoft working on Teams feature to keep unauthorized bots at bay
Microsoft plans to add a new Teams feature that lets meeting admins identify and control third-party bots before they join. According to the Microsoft 365 Roadmap, the feature is scheduled to begin rol… #hackernews#microsoft#news
A profile of Simile, which offers "agentic twins" modeled on real people to provide answers for polling and market research for companies like CVS and Gallup (Belle Lin/Wall Street Journal)
Engineer receives $30,000 for exposing a vulnerability affecting 7,000 robot vacuum cleaners — tinkerer just wanted to drive his robot vacuum with a PS5 controller A software engineer gets access...
Digital footprints lead to the heart of federal surveillance. After detecting abnormal logs, FBI confirms hackers breached a system holding sensitive pen register returns. The incident uncovers vulnerability of even the most guarded law enforcement data to sophisticated external exploitation.
"the company announced Claude Marketplace, a new offering that lets enterprises with an existing Anthropic spend commitment apply part of it toward tools and applications powered by Anthropic's Claude models but made and offered by external partners" venturebeat.com/technology/a...
"We show that large language models can be used to perform at-scale deanonymization. With full Internet access, our agent can re-identify Hacker News users and Anthropic Interviewer participants at high precision, given pseudonymous online profiles and conversations alone." arxiv.org/html/2602.16...
I tried Gemini, ChatGPT, and Copilot for a month and my findings will shock you Don’t buy the hype. I went into this month-long experiment convinced that the differences between Gemini, ChatGPT, ...
It's mostly reflective of tech industry executives or others trying to sell something, rather than researchers, but I must sadly admit that some researchers act this way too.
OpenAI released GPT-5.4, with a 1,000,000-token context window (up from 272,000), about 20% performance gain over GPT-5.2, native computer interaction at 75% accuracy, dynamic tool loading reducing token use 47%, and slightly higher Pro pricing.
Anthropic released Claude Cowork, an experimental desktop agent in the Claude app for Windows and macOS (research preview on paid plans) that accesses local files and connectors, executes scheduled tasks, produces step-by-step plans, and handles xlsx, pptx, docx and pdf.
"Dr. Amodei [said] Anthropic failed to win over the Pentagon because it did not give 'dictator-style praise' to the Trump administration the way he said Mr. Altman was willing to do."
Microsoft spots ClickFix campaign getting users to self-pwn on Windows Terminal
Crooks tweak familiar copy-paste ruse so that victims run malicious commands themselves
A new twist on the long-running ClickFix scam is now tricking Windows users into launching Windows T… #hackernews#microsoft#news
Bend over Windoze users, Micro$uck has even more ridiculous nonsense heading your way!
Microsoft's Secure Boot certificates expire in June 2026, but older PCs may never get the fix share.google/KKDLaYqJ0LQ7...
I guess Anthropic didn’t want to be compared to IBM. IBM’s German subsidiary used its punch cards and other business equipment to facilitate the roundup of Jews and the implementation of the Holocaust.
Security researchers at Black Kite have warned that a vast, largely hidden layer of third‑party breaches went unreported last year, exposing tens of thousands of supply‑chain firms and more than 433 million people.
🔎 Google said it will continue offering Anthropic’s artificial intelligence technology for clients, excluding for defense work, a day after Microsoft issued a similar statement to customers. 💯 www.cnbc.com/2026/03/06/g...#Anthropic#Google#Claude
Anthropic's AI helped the US military strike over 1,000 targets in Iran in one day. Anthropic fought the Pentagon over how Claude would be used. They lost. The company that builds the tool does not always control what the tool becomes.
www.macworld.com/article/2854... "The MacBook Neo is a new laptop from Apple that is sure to be a big hit because it comes in at a much lower price than the MacBook Air, starting at $599/£599 compared to the MacBook Air’s $1,099/£1,099." BUDGET LAPTOP MEANS UNDER 400€!!!!
Microsoft becomes first company to say it is not 'abandoning' Anthropic; company says: Our lawyers have studied that ... | - The Times of India share.google/Gh1SwWNsMx6u...
Iran-linked APT targets US critical sectors with new backdoors
An Iran-linked hacking group has been active inside the networks of several US organizations since early February, raising concerns that the activity could precede broader cyber operations connected to escalating geop… #hackernews#news
Low paid workers in the global south are using AI translation to generate Wikipedia articles that, surprising no one, include errors and hallucinations.
🛑The new Gemini Canvas gives you a side-by-side editor to refine your writing and code in real-time. Build a game or a 1,000-word launch plan with ease.
I've used a Raspberry Pi for 10 years, and this easy project started it all An off-the-shelf kit and about two hours are all you need. Getting started with a Raspberry PI can feel daunting, esp...
My guess with LLMs is there will be a major market correction, then a slew of bankruptcies and investor lawsuits, followed by industry-wide consolidation/acquisitions, then a wave of anti-trust enforcement, and eventually it will settle into some sort of oligopoly.
"A similar study found a 16% fall in employment in jobs exposed to AI among workers aged 22 to 25."
Anthropic just mapped out which jobs AI could potentially replace. A ‘Great Recession for white-collar workers’ is absolutely possible share.google/M9eG1zkXZGQA...
Are you still using the Proton Mail? Well if the fact that the guy behind it is a fascist isn't enough to get you to stop, maybe the part where it is selling y'all out to the cops will: www.wired.com/story/cbp-us...
“Because many IT worker campaigns rely on the abuse of legitimate access, Microsoft advises organizations to treat these schemes and similar activity as insider risks.”
Sources: Palmer Luckey's ModRetro, which wants to sell updated versions of 1990s consoles like the Nintendo 64, is in talks to raise funding at a $1B valuation (Financial Times)
Mozilla says Claude AI uncovered over 100 Firefox bugs in just two weeks, including 14 high-severity flaws: Mozilla is now working with Anthropic's Frontier Red Team to identify and patch potentially dangerous security vulnerabilities in Firefox. According to Mozilla, the AI company approached them…
AI ethics brings up "the same kinds of questions that people have been asking for centuries. What does it mean to be a human? What does it mean to have responsibilities to others? What kind of responsibilities do we have for the things that we create?" www.ie.edu/insights/vid...
I tore apart the most common Linux malware in a sandbox, and it uses layer after layer of tricks to survive It uses some of the oldest tricks in the book. There's a common misconception that Li...
Cannot believe this meetup wasted the opportunity to get a white claw sponsorship and make its tagline "ain't no laws when you're making claws" www.theverge.com/ai-artificia...
Caitlin Kalinowski, OpenAI's head of hardware and robotic engineering, resigns citing concerns over domestic surveillance and lethal autonomous weapons systems (Sharon Goldman/Fortune)
Had a long, candid talk with students about vibe coding. I am shocked that instructors are not disclosing risks up and down the stack. What a way to get blackballed from your first pro job by injecting malware into a product/analysis!
Delayed "in order to 'focus on work that is a higher priority for more users right now,' such as work on aspects like intelligence, personality, and making the chatbot 'more proactive'
Or maybe unable to guarantee guardrails against depicting children and real people
Join EFF Exective Director Cindy Cohn and @404media.co Cofounder Jason Koebler for an in depth conversation about Cindy's new book "Privacy's Defender: My Thirty-Year Fight Against Digital Surveillance" live on March 19 at 11 am PT eff.org/livestream-...
Why the US government labelled Anthropic a ‘supply chain risk’: A timeline: The U.S. government designated artificial intelligence company Anthropic as a "supply chain risk" due to concerns about its reliance on foreign-made semiconductors, particularly those manufact… https://ranked.news/284493?u=b
For International Women’s Day, we’re highlighting the contributions of just a few EFF Award recipients from the last decade, whose work to protect privacy, speech, and creativity online has had a global impact.
Codex for Open Source Codex for Open Source Anthropic announced six months of free Claude Max for maintainers of popular open source projects (5,000+ stars or 1M+ NPM downloads) on 27th February . ...
EC-Council Expands AI Certification Portfolio to Strengthen U.S. AI Workforce Readiness and Security
EC-Council, creator of the world-renowned Certified Ethical Hacker (CEH) credential and a global leader in applied cybersecurity education, today launched its Enterprise AI Creden… #hackernews#news
“Caitlin Kalinowski, who had been leading hardware and robotic engineering teams at OpenAI since November 2024, announced she has left the company.” fortune.com/2026/03/07/o...
Gemini won't load, responses cut off, or Workspace AI is greyed out? 9 fixes for the most common Gemini problems. https://techsifted.com/guides/gemini-ai-not-working-fixes/ #Gemini #AI
OpenAI delays adult mode : We still believe in the principle of treating adults like adults, but getting the experience right will take more time. A good and fairly obvious call, this was a distra...
www.wired.com/story/coruna...
A refined iPhone hacking toolkit called Coruna has infected tens of thousands of phones. Experts believe it may have originated from tools built for the US government, but it later spread to Russian spies and cybercriminals who used it to steal data and cryptocurrency.
OpenAI head of robotics just resigned over company deal with the Pentagon saying…
“Surveillance of Americans without judicial oversight and lethal autonomy without human authorization are lines that deserved more deliberation than they got”
I've just used AI to generate some legal documents (simple ones) and ask questions about taxes, so far it worked wonderful. Gemini and Claude were on par.
Not sure to which one to continue subscribing too, probably Gemini on workspace individual because it's cheaper and comes with additional stuff.
@rbreich.bsky.social Trump fires US security experts, installs loyal yes-men, pardons rich criminals, and sells US top secret computer components to China.. Big surprise.... www.wsj.com/politics/nat...
Gemini 3.1 Pro just hit 77% on ARC-AGI-2. A year ago that benchmark was the 'AI can't do this' wall. Now it's a score to brag about. The goalposts move faster than anyone predicted
Leading the Future, a pro-AI PAC backed by Palantir-cofounder Joe Lonsdale, hit pro-regulation Democrat Alex Bores with attack ads over Bores' work for Palantir (Nancy Scola/Politico)
Do not commit any crimes and post them on social media. Even Proton the privacy email company will give up your info when asked by the government. #resist#privacy#proton#activism#noprivacy
TCL’s TAB A1 Plus focuses on immersive entertainment!
With a 12.2-inch 120Hz display, quad speakers, Gemini integration, and premium materials, it’s built for work & play.
#TechRadar 'What begins as a phone call from 'IT support' ends with a fully instrumented network compromise': This fake tech support scam tricks employees into infecting their own company devices https://techrad.ar/SPTK#Security#Pro
![Media from Michael Georg Speller [THE GREY-APE]](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:422uaueaquotbw7ie6th23qv/bafkreiezmeqhjxrb5vjwd2qnwukeklkdpi4p5ker7h5vdtqmofhgt5p4dm)
![Media from Michael Georg Speller [THE GREY-APE]](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:422uaueaquotbw7ie6th23qv/bafkreicwgw3womkbi5fhj6alcvohx6ecciqzw7fozsgoggmec6svnqeuo4)
![Media from Michael Georg Speller [THE GREY-APE]](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:422uaueaquotbw7ie6th23qv/bafkreifdgnmwjzzd3d4fbzwdrtm5ykljusojlogm4djh4hr4boe2xr6vbe)
