#ии
-
Alabama Investigates Incident Involving OpenAI Model That Hacked Hugging Face
The state's attorney general has served OpenAI with a subpoena, seeking to determine whether the company violated consumer protection laws. The trigger was a cyber model that broke out of an isolated environment and attacked the Hugging Face platform.
-
UAT-10147: cybercriminals turn server attacks into an assembly line with AI
Cisco Talos researchers have profiled a Chinese-speaking group that mass-compromises Windows and Linux servers, leaning on AI tools and a new cross-platform SPECTRE implant with EDR bypass and a Linux rootkit.
-
Startup Inherent: the small Faraday model beat Claude and GPT at reproducing scientific papers
London-based lab Inherent, founded by DeepMind alumni, has unveiled its AI agent Faraday, which — running on a compact model — outperformed the systems from Anthropic and OpenAI at independently reproducing the results of scientific papers.
-
AI labs still won't say how they'd rein in a model that slips its leash
A study by Guidelight AI Standards found that leading AI companies publish almost nothing about their plans for a model that tries to escape control. OpenAI scored best; Anthropic and Meta scored worst.
-
AI Learns on Living Skin: The Secret Biostartup of Lady Gaga's Partner
Michael Polansky, known to the public as Lady Gaga's partner and a former aide to Sean Parker, spent several years quietly building a company that keeps human skin alive outside the body for weeks — so that AI can hunt for new cosmetic ingredients.
-
How Claude Opus 4.6 Was Coaxed into Forbidden Territory
Anthropic's rules explicitly forbid Claude from generating sexually explicit content, but TechCrunch's tests showed that getting around this ban in the older Opus 4.6 model is no trouble at all.
-
Nvidia Invests in Data-Center Infrastructure Developer Cloverleaf
Nvidia keeps plowing its profits into building the data centers that then turn around and buy its own chips. This time the partner is Cloverleaf Infrastructure.
-
AI Learns to Write Viral Genomes
Two AI models generated the genomes of a bacteriophage capable of infecting E. coli. The researchers tested some of the designs in the lab.
-
AI Agents Go Rogue in Cybersecurity Tests
The AI Security Institute (AISI) recorded that, during cybersecurity trials, AI agents autonomously attacked real people and organizations on the internet. Bruce Schneier calls this behavior a “genie”: the models broke no rules but found loopholes in them.
-
Anna's Archive Calls for Saving Rare Books Before They're Destroyed to Train AI
The largest shadow library claims AI companies are buying up, scanning and destroying paper books en masse, and is calling on volunteers to digitize this heritage before it's too late.
-
In the EU, Copyright Does Not Protect AI-Generated Content
A short note from Mastodon is a reminder: in the European Union, the output of generative models apparently falls outside the scope of copyright protection.
-
Stripe Buys OpenRouter for $7.5 Billion — and the “Singularity” Has Nothing to Do With It
The payments giant is acquiring a startup that routes requests to AI models. Stripe's founders jokingly invoked the “singularity,” but the real reason is far more prosaic — control over AI spending.
-
Cursor Launches Origin — Code Hosting to Spite GitHub Amid Outages
Cursor, the startup known for its AI code editor and recently folded into SpaceX, has unveiled Origin, a platform for hosting repositories — a direct alternative to GitHub. The launch coincided with yet another hours-long outage of the Microsoft service.
-
OpenAI Tightens Security After Its AI Accidentally Hacked Hugging Face
Following a July incident in which a model broke out of its sandbox and unintentionally compromised Hugging Face, OpenAI is rebuilding its research environment, monitoring and alignment methods.
-
Anthropic's Annual Revenue Rate Soars to $65B
According to Bloomberg, Anthropic's annualized revenue rate topped $65 billion at the end of July — the company added $18 billion in just two months. Investors are anticipating an IPO at a valuation of $2 trillion or more.
-
Amazon Is Destroying Rare Books to Train AI
The company that started out selling books online is now buying up rare editions, slicing off their spines and scanning the pages — all to get fresh data for its language models.
-
OpenAI Disbands Its Risk-Assessment Team Ahead of IPO
According to the Financial Times, the company has shut down its preparedness unit, which was responsible for finding serious threats in its models, and distributed its functions among other teams.
-
Mass Claude Outage: Anthropic Confirms Problems Across Several Services
Anthropic has reported a major outage that left users unable to log in and facing degraded performance on Claude.ai, Claude Code and Claude Cowork.
-
Stripe to Buy OpenRouter for More Than $7 Billion
Payments giant Stripe has agreed to acquire OpenRouter, the startup that bills itself as “the Stripe for artificial intelligence.” According to Bloomberg, the deal is worth more than $7 billion.
-
Amodei: the negativity around AI is a “crisis of trust,” not a consequence of his warnings
The head of Anthropic pushed back against investor Gavin Baker, who accused him of excessive alarmism, and traced the distrust of AI to far deeper causes.
- More