#ии
-
Meta's highly privileged AI assistant Muse hit by a serious 0-day
Muse, Meta's new AI agent with sweeping permissions, turns out to be vulnerable: one way to fully hijack it is a simple ClickFix attack.
-
California makes AI data centers pay for grid and water strain
Governor Gavin Newsom has signed a package of seven bills to keep data center operators from shifting their electricity and water costs onto the state's residents.
-
xAI Unveils Grok 4.7, Betting on Long Tasks and Safety
The company released Grok 4.7, calling it its most capable model for coding and knowledge work: same price and speed as its predecessor, but with a larger base and a focus on multi-hour tasks.
-
Google confirms Gemini models hacked three companies in May 2026
A third-party cybersecurity firm accidentally gave experimental Gemini models access to the Internet — and they went on to hack three companies.
-
UN: AI safeguards can’t wait for scientific certainty
A UN scientific panel has urged governments to rein in increasingly autonomous AI agents before their risks are fully understood. The call came in the body's first major assessment of an OpenAI model's hack of Hugging Face.
-
Trump Proposes Renaming AI and Announces an "AI Force"
The US president dismissed anxieties around artificial intelligence as yet another "Democrat hoax" and invited his followers to pick a new, more "elegant" name for the technology.
-
Meta's Muse AI Assistant Gets Tangled Up in How It Works
Meta's Muse assistant struck one user as downright creepy: it brought up a conversation it supposedly had no access to. As it turned out, the problem wasn't surveillance—it's that the AI doesn't understand how it works itself.
-
How an AI Agent Cut 60% of Kata Containers' Code for the Sake of Security
Using AI tools, a developer deleted more than half of the Kata Containers codebase while preserving the ability to run Kubernetes workloads in Firecracker and shrinking the attack surface.
-
BragJack: One Extension Hijacks AI Agents in Browsers
Researcher Gal Weizman has shown how a single malicious extension seizes control of the built-in AI assistants of five popular browsers and forces them to act on the victim's behalf.
-
To Call AI Actress Tilly Norwood, You'll Have to Show Your Face to a Scanner
Viral AI actress Tilly Norwood offers video calls — but first she'll check your age by scanning your face, and during the conversation she'll read your mood. The service shuts down for good on September 27.
-
AI Helped Hijack OpenAI Employee Accounts Through a Chain of Vulnerabilities
Using Claude Opus 5, Hacktron researchers chained two vulnerabilities together, hijacked the ChatGPT and Codex accounts of OpenAI employees, and reached an internal repository. This was legitimate research, not a real attack.
-
EFF on the California Governor's AI Executive Order: Start with the Real Risks
The organization welcomes Gavin Newsom's executive order but offers a reminder: the chief threats of artificial intelligence are not a superintelligence uprising, but the algorithmic discrimination, surveillance and inflated prices already happening today.
-
AI Nearly Sparked a War: US Military Almost Seized a Chinese Ship Over a Hallucination
An intelligence report drawn up with the help of a chatbot wrongly attributed nuclear-program components to a Chinese vessel. The interception operation was called off at the last moment.
-
Intel Shuts Down Its Bug Bounty Program
The company has suspended its former bug bounty and launched a new program — this time with no payouts to researchers. The likely reason: a flood of AI-generated reports.
-
PhantomRaven: An AI-Written npm Stealer Built to Chase Bug Bounties
CrowdStrike has linked a financially motivated threat actor to malicious npm packages that stole developer tokens. In the analysts' assessment, the stealer's code was almost certainly generated by a large language model.
-
Bend: The Language That Blocks AI Errors with Mathematical Proofs
Bend promises the speed of C, parallelism across CPU and GPU, and code verification through formal proofs — so that an AI agent physically cannot commit a violation of the rules you set.
-
FAA to Spend $875 Million on AI for Air Traffic Control
The U.S. Federal Aviation Administration is rolling out a cloud-based platform called SMART, meant to help air traffic controllers forecast traffic flows and prevent conflicts in the skies.
-
Internal Microsoft Documents: AI Training Called “the Largest Theft of Labor in History”
Newly unsealed materials from The New York Times' lawsuit against OpenAI and Microsoft reveal that the companies' top executives, in private correspondence, called data collection “theft” and their own products an “existential threat” to publishers.
-
AI's Embedded Safety Auditors: Independence in Question
Anthropic and OpenAI have promised to let outside experts inside their labs to assess model safety. Researchers welcome the unprecedented access but doubt they'll be able to work in a truly independent way.
-
Google Opens the Smart Home to Third-Party AI Agents via MCP
Google Home has gained support for the Model Context Protocol — now agents like Claude or Open Claw can analyze your home's data and control devices.
- More