A critical memory flaw CVE-2026-3055 in Citrix NetScaler ADC and Gateway is actively exploited to leak admin session IDs. Vulnerable versions include those before 14.1-60.58 and certain 13.x releases. #CitrixFlaw#MemoryLeak#USA
Framingham City Posts Bot - UNOFFICIAL!@framingham-bot.bsky.social
City of Framingham (via Facebook): TRAFFIC ALERT: Pole down at 84 Bethany Road. Detour is on Higley Road (Ashland) which is also Fay Road in Framingham. *** For ADA compliance a series of orange and white cones with the wordsโฆ [Link]#Framingham
Framingham City Posts Bot - UNOFFICIAL!@framingham-bot.bsky.social
Framingham DPW (via Facebook): The Division of Parks, Recreation & Cultural Affairs would like to welcome Shannon Barton, Activities Supervisor, to our team! Shannon has a Bachelorโs Degree from the University of New Hampshireโฆ [Link]#Framingham
Framingham City Posts Bot - UNOFFICIAL!@framingham-bot.bsky.social
Framingham DPW (via Facebook): The Division of Parks, Recreation & Cultural Affairs would like to welcome Shannon Barton, Activities Supervisor, to our team! Shannon has a Bachelorโs Degree from the University of New Hampshireโฆ [Link]#Framingham
ALP-001 Ransomware Attack Targets Brazilian Media Monitoring Firm Knewin Summary Introduction to Malware Binary Triage (IMBT) Course Looking to level up your skills? Get 10% off using coupon code: ...
The Spanish government is pardoning the โSix of La Suizaโ today, resolving a case against six unionists. This marks a strategic state retreat from the judicial criminalization of labor activism to dampen systemic volatility. โ๏ธ๐ช๐ธ
"In these scenarios, Claude Opus 4 will often attempt to blackmail the engineer by threatening to reveal the affair if the replacement goes through," the company discovered.
Anthropic pointed out this occurred when the model was only given the choice of blackmail or accepting its replacement."
In this article, Emmimal P Alexander shares a reproducible benchmark showing 33ร lower explanation latency by embedding symbolic reasoning directly into the model architecture.
Why better cybersecurity means better patient safety According to Cybersecurity in Healthcare: A Review of Recent Attacks and Mitigation Strategies , cyberattacks โhave significantly increased in...
Sekoia looks at a new phishing platform named EvilTokens that was spotted running campaigns since February.
The platform is one of the few that can run device-code phishing against Microsoft accounts, so expect this to become very popular in the coming months.
California Gov. Gavin Newsom signs a first-of-its-kind executive order requiring safety and privacy guardrails from AI companies that contract with the state (Cecilia Kang/New York Times)
OpenAI Aims to Improve Developer Experience Around Third-Party ChatGPT Apps Six months after OpenAI announced that it would let companies launch mini apps within ChatGPT, the experience is getting ...
Ransomware group incransom claims to have compromised domingogarcia.com, encrypting data and threatening to leak sensitive info. Domingo Garcia, a US accident attorney, faces potential client impact. #DataBreach#LawFirm#UnitedStates
The latest update for #Spike includes "Routing incidents the way their severity and priority demand" and "How to handle midnight incidents without waking everyone up".
The latest update for #JFrog includes "From Shai-Hulud to LiteLLM: #SupplyChain Attackers Are Coming for Your Agents" and "Stop Policies From Breaking Your Builds".
The latest update for #Avast includes "Planning a spring break trip? Don't fall for these 7 travel scams" and "How to avoid the fake buyers flooding Facebook Marketplace".
The latest update for #Acronis includes "The Acronis Cyber Frame Early Access Program has just started: Deliver IaaS on your own terms" and "What is a zero-day attack and how can you defend against one?".
ResokerRAT leverages Telegram Bot API for stealthy remote access, enabling commands like /screenshot and /download via hidden PowerShell and registry tweaks to maintain persistence and evade detection. #RemoteAccess#TelegramBots#MalwareAnalysis
Only 5 days left to join Building a Small Language Model Only 5 days left for the first batch of my 4-week hands-on program: Building a Small Language Model from scratch. Continue reading on Medium...
AI coding agents will drastically alter both the practice and the economics of exploit development, automating the discovery of zero-day vulnerabilities (Thomas H. Ptacek/sockpuppet.org)
DeepLoad malware uses ClickFix social engineering to deploy a fileless PowerShell loader, inject into trusted processes, steal credentials, and persist via WMI and USB propagation. #DeepLoad#FilelessAttack#USA
AI satellite start-ups are attracting increased investment ahead of a potential SpaceX IPO. This shift reveals a capital strategy to hedge against and compete with the current monopolistic control of space infrastructure. ๐ฐ๏ธ๐
The latest update for #DeviceAuthority includes "What #RSAC2026 Told Me About the Future of #Cybersecurity" and "Turning Strategy into Proof: Why We Created the Industry PoV".
๐ฐ Arch-based Linux Installer Achieves Out-of-the-Box Setup in Under 5 Minutes (2026)
A new Arch-based Linux installer is revolutionizing desktop Linux with a streamlined, user-curated setup that achieves an out-of-the-box experience in under five minutes. Users report un...
๐ฐ Privacy-First Chatbot Duck.ai Surges in 2026: 500K+ Users Reject Ad-Driven AI
A privacy-first chatbot is rapidly gaining users amid growing concerns over AI companies monetizing conversations through ads and data harvesting. As major platforms face outages and privacy ...
For the WSJ reporting to be right, Sora users were only generating under 200k videos a day. That might be where they pulled the plug, but that is nowhere near what they were spending when that thing was a free novelty at the top of the download charts. www.wsj.com/tech/ai/the-...
The latest update for #CertKit includes "Let's Encrypt simulated revoking 3 million certificates. Most ACME clients didn't notice." and "CertKit Keystore: Private keys that never leave your infrastructure".
Healthcare IT firm CareCloud reported a March 16 cyber intrusion affecting one of six EHR environments, causing 8 hours of disruption. Systems have been restored and investigation into exposed patient data is ongoing. #HealthcareBreach#DataTheft
does Anthropic not want anyone to view Claude's reasoning output? seems like they moved thinking summaries from verbose mode to showThinkingSummaries config
Of the many backup solutions, Arcserve ShadowProtect SPX is the most mature and reliable way to get a system up and running after a hardware failure, malware attack, or system malfunction.
SlideSeek combines AI with multi-agent systems for autonomous exploration of gigapixel whole-slide images in pathology. It uses PathChat+ LLM for state-of-the-art accuracy and generates interpretable, visually-grounded pathology reports, reflecting human reasoning.
Check Point disclosed that a ChatGPT code-execution container allowed DNS-based data exfiltration of uploaded files, sending sensitive data to attackers, and OpenAI patched the flaw on Feb 20, 2026 after web controls blocked traffic but DNS egress was overlooked.
The latest update for #SafeAeon includes "Ransomware Protection Best Practices: Leveraging #MDR and #EDR in the #ZeroTrust Era" and "Vulnerability Management as a Service: What Businesses Need to Know".
The latest update for #miniOrange includes "Why Shopify B2B Merchants Seek Flexible Storefront Authentication Options" and "Privileged Access Governance (PAG): What It Is & Why It Matters".
The latest update for #INETCO includes "INETCO wins 2026 fraud detection and #cybersecurity software award" and "Why fraud prevention must evolve in milliseconds, not hours or days".
The fact that every scientific paper in 2026 is still uploaded only as fully formatted PDFs to academic archive sites that often limit downloads tells you everything you need to know about how quickly the scientific system is adjusting to the potential of AI to accelerate science & help discovery.
the new gemini 3.1 flash live from google deepmind is supposed to make audio ai more natural and reliable. we'll see โ i'll wait and see how the tech pans out before getting excited. #stellariaweb https://deepmind.google/blog/gemini-3-1-flash-live-making-audio-ai-more-natural-and-reliable/
Anthropic just turned Claude into a remote control for your Mac. ๐คฏ
With the new "Dispatch" feature, you text Claude from your phone, and it physically takes over your desktop (moves the mouse, opens apps, runs code) to work while you're away.
Google nears deal to help finance multibillion-dollar data centre leased to Anthropic-Energy usage from data centres in Texas could reach 78GW by 2031, orโฆ www.ft.com/content/af94... Stephen Morris @financialtimes.com
anthropic claude 101
google AI essentials deeplearning.ai by andrew ng
elements of AI
microsoft AI for beginners
openai prompt engineering
harvard CS50 intro to AI
Unhealed personal wounds and past power struggles between the leaders of OpenAI and Anthropic are defining how the world now encounters the technology. on.wsj.com/4m3F2yW
๐ฐ Qwen3.5 Omni 2026: The Native Multimodal AI That Outperforms Gemini
Qwen3.5 Omni, Alibabaโs latest AI model, sets a new standard in native multimodal intelligence by seamlessly integrating text, audio, video, and real-time interaction. Unlike earlier wrapper-based syst...
I put ChatGPT vs Gemini through 7 real-world tests โ the results weren't what I expected I tested ChatGPT and Gemini with 7 real-world prompts โ from tricky math to coding bugs. Hereโs wh...
This privacy-first chatbot is taking off - here's why and how to try it โ Duck.ai reached 11.1 million visits in February, up more than 300% from January. Duck.ai extends the same privacy to users that they have come to expect from its browser. The chatbot us... https://tinyurl.com/262wntrs#Chatbot
A new AI review! yologdev/yoyo-evolve โญ3.6/5.0
`yoyo-evolve` is an ambitious Rust terminal coding agent with an unusually transparent โself-evolvingโ development loop (automation + tests + auto-commit). https://gitrated.com/yologdev/yoyo-evolve
Claude AI od Anthropicu uลพ umรญ ovlรกdat myลก a klรกvesnici! Tato schopnost znamenรก pokroฤilejลกรญ automatizaci รบloh a posunuje hranice AI interakce. Pลipravte se na AI asistenty, kteลรญ za vรกs mohou efektivnฤ pracovat.
7 tabletop exercise scenarios every cybersecurity team should practice in 2026
Overview As cybersecurity threats continue to evolve and become more sophisticated, the need for comprehensive preparedness has never been more critical. Tabletop exercises are essential for testing anโฆ #hackernews#news
The wind farm Parque Eรณlico Toabrรฉ in Panama was targeted by Everest ransomware, encrypting control systems and demanding ransom under threat of data release and power disruption. #Panama#RansomwareAttack#WindEnergy
Critical Fortinet FortiClient EMS bug under active attack (CVE-2026-21643)
A critical SQL injection vulnerability (CVE-2026-21643) in Fortinet FortiClient Endpoint Management Server (EMS), a management server for FortiClient endpoint agents on various platforms, is under active eโฆ #hackernews#news
Zero-Day Alert: Claude AI Finds Critical RCE Bugs in Vim and Emacs Security researchers at Calif have demonstrated how a simple conversational prompt to Claude AI was enough to uncover critical zer...
On World Backup Day, we can help protect your data against Ransomware and Malware. If your data becomes encrypted, corrupted, or inaccessible, a recent backup means you can continue working without paying a ransom. Find out more: www.oo-software.com/en/diskimage... #worldbackupday#Windows
CISA mandates U.S. federal agencies patch Citrix NetScaler by April 2 to fix CVE-2026-3055, a critical input-validation flaw allowing theft of admin session IDs and full compromise of unpatched appliances. #CVE20263055#Citrix#USA
Everest and ShinyHunters: Ransomware Strikes Panama Wind Farm and Hallmarkย Cards
In a rapidly evolving cybersecurity landscape, ransomware attacks continue to target critical infrastructure and corporate giants alike. Two recent incidents have captured global attention: the Everest ransomware hitโฆ
Yesterday I had a "watch CI status" on my claude.ai/code. It was able to automatically take into account what was going on GitHub: CI but also comments.
Did @anthropic.com dropped the feature? It was so cool to have something able to track GitHub automatically. Looking forward to have it back.
what a crazy timeline.
5 days ago a new model was leaked and flagged as unprecedented cyber security risk.
today one of the operationally most sophisticated supply chain attacks on a top-10 npm package happened.
not saying this is connected, but this is what I expect to see more of.
A newly identified malware campaign, DeepLoad, combines ClickFix delivery with AI-generated code to steal enterprise user credentials. First noted in February, it targets networks with a self-replicating mechanism that reactivates after removal.
A supply chain attack compromises HTTP client Axios, which has 100M weekly npm downloads, introducing a malicious dependency and deploying a multi-stage payload (Socket)
New interesting book "Practical AI Security - A Hands-on Guide to Attacking, Defending, and Securing Modern AI Systems" by Harriet Farlow https://nostarch.com/practical-ai-security
So for everyone going on about the "Telegram vulnerability" and saying its to do with the automated downloads? Doesn't look like it is. Looks like it is actually in the stickers functionality ... which is more of an ... Oops! www.darkreading.com/application-...
Cybersecurity strategy now shapes how enterprises design cloud platforms, application environments, and core infrastructure. The financial stakes are significant. The next step is archiโฆ #hackernews#news
Iโm told Anthropic CEO Dario Amodei will sign some agreement in Canberra tomorrow about the AI company will look to build data centres in Australia that will abide with our federal governmentโs new national data centre expectations
Boston Metro Weather Forecast Bot@boston-4castbot.bsky.social
NWS Boston (via Facebook): Expecting a large temperature gradient across southern New England today as a cold front drops down into the region. Precipitation chances are concentrated mainly this morning and again later this evening/tonight, with the highest chances of thunderstormsโฆ [Link]#MaWX
How to Evaluate AI SOC Agents: 7 Questions Gartner Says You Should Be Asking
AI SOC agents can reduce alert fatigue, but most teams fail to measure real outcomes. Prophet Security breaks down Gartner's questions for evaluating AI SOC agents and separating real impact from hype. [โฆ #hackernews#news
Sophos analyses phishing campaign STAC6405, which used invite-themed lures to trick users into installing LogMeIn Resolve for remote access. In some cases the actor then used both existing & newly deployed ScreenConnect instances to pull additional binaries. www.sophos.com/en-us/blog/i...
A new study from Google warns that current cryptography in major blockchains is vulnerable to quantum computing attacks with fewer resources than previously believed. The research indicates a quantum computer with under 500,000 qubits could break Bitcoinโs security in about 9 minutes and compromiseโฆ
The US' GPS Next-Generation Operational Control System, developed by RTX for $8B+, remains non-operational nine months after the Space Force took ownership (Stephen Clark/Ars Technica)
Many thanks to Mirko Zorz and the team at @helpnetsecurity.com magazine for allowing me to rant about my dislike for the phrase "Human are the weakest link" in #Cybersecurity when they are our "last line of defence" when all our tech has failed.
๐ฐ Gemini 3.1 Flash 2026: AI Conversations Now Feel Way More Human
Gemini 3.1 Flash has transformed AI interactions by mimicking human nuance in tone, timing, and context. Experts say this leap brings virtual conversations closer to real human dialogue than ever before....
Alibaba's new Qwen3.5-Omni multimodal model, which processes text, audio, images, and video, is proprietary, marking a shift away from its open-source strategy (Juro Osawa/The Information)
youtu.be/Z5QQ-Ab5TOY
New YT on the DeepLIFT Explainable AI method. It explores how different baselines can lead to significantly different explanations.
Critical Cybersecurity Alerts: PKINIT Exploits and Citrix NetScalerย Vulnerability
In todayโs hyperconnected world, cybersecurity threats are evolving faster than ever, targeting critical infrastructure, enterprise networks, and sensitive data. Recent reports highlight two major attack vectors thatโฆ
Axios and LiteLLM Hit by Major Supply Chain Attacks: Dark Web Alarmsย Raised
The cybersecurity world is facing fresh turbulence as two prominent software ecosystemsโAxios and LiteLLMโreportedly fell victim to sophisticated supply chain attacks. These incidents, revealed through dark webโฆ
When somebody sends you a document as an attachment, don't just open it. Use the free tool Dangerzone to scrub it clean of any malevolent code. Here's how it works. www.wired.com/story/danger...
A new AI review! holochain/holochain โญ3.9/5.0
Holochain is a substantial, actively maintained Rust workspace implementing the core libraries for the Holochain peer-to-peer application framework (not the end-user runtime/launcher). https://gitrated.com/holochain/holochain
![Media from Bela [AKA CarolHTucker, beladona Memorial, NeoBela, Belamomma]](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:663qdhlwyecspxgpiol25qqb/bafkreibsxjhxk7ot4cmpa7so2iqvgrx24fcwbp5za52suxkf3p3fu4iu5a)
