#искусственный-интеллект
-
US Lawmakers Draft an AI “Kill Switch” Bill
Congressmen Ted Lieu and Nathaniel Moran are introducing a bill that would let the Department of Homeland Security order companies to shut down or slow powerful AI systems.
-
RefluXFS: a Local Privilege Escalation to root in XFS, Found with AI
Qualys has disclosed CVE-2026-64600 — a race condition in the copy-on-write subsystem of the XFS file system that lets an ordinary user gain root privileges. The vulnerability has existed since kernel 4.11 and potentially affects more than 16 million systems.
-
Google Unveils Gemini 3.6 Flash and a Cybersecurity Model, While Gemini 4 Is Already in Training
The company has released three new models, including the first version of Gemini built for cybersecurity tasks — but the promised Gemini 3.5 Pro is once again nowhere among them.
-
The Suno Breach: Data of 55 Million Users at Risk
The Have I Been Pwned service has gauged the scale of last year's attack on the music generator Suno: the intruder made off with the personal data of more than 55 million people. The company still hasn't disclosed the incident publicly.
-
Court Approves Record Anthropic Payout to Authors — $1.5 Billion
A federal court has given final approval to a settlement over copyright infringement: Anthropic will pay roughly $3,000 for each of about 500,000 works. But the industry's central question remains open.
-
JadePuffer arms itself with the EncForge ransomware to attack AI model data
The autonomous AI agent JadePuffer has acquired specialized malware called EncForge, which encrypts training datasets, vector databases and model checkpoints. Sysdig estimates that the damage to a single model could run from $75,000 to $500,000.
-
Fable Language Model “Disproves” the Jacobian Conjecture — but Mathematicians Are Cautious
A mathematician going by the handle levent announced on the social network X that the famous Jacobian conjecture is false: the counterexample was built with the help of the AI model Fable. Verification of the result is still ongoing.
-
AI Tips Made People Three Times Less Accurate but Twice as Confident
Researchers from French and Italian universities have found that access to AI advice all but strips people of their ability to admit “I don't know” — and sharply lowers the accuracy of their answers to boot.
-
Moonshot AI Pauses New Subscriptions Amid Demand for Kimi K3
Chinese company Moonshot AI has temporarily halted new sign-ups: demand for its Kimi K3 model has run up against the limits of its computing capacity.
-
Alibaba Prepares to Release Open Model Qwen3.8
The Qwen developers have announced the launch of Qwen3.8, whose weights will soon be available in open access. The news drew notable interest on Hacker News.
-
How AI Euphoria Is Undermining Corporate Governance
Columnist Nick Suresh has gathered anecdotes from anonymous sources about how the hype around AI distorts decision-making at large companies. The piece was reposted and commented on by developer Simon Willison.
-
Author Dave Eggers told OpenAI staff that ChatGPT is “silencing an entire generation”
Author Dave Eggers spoke to OpenAI employees at Sam Altman's invitation and declared that ChatGPT is suppressing the creative voice of an entire generation.
-
GPT-5.6 Suggested a Solution to a 30-Year Convex Optimization Problem
According to Reddit users, the language model GPT-5.6, guided by a specially crafted prompt, helped close a question in convex optimization that had been open for thirty years. The news sparked a lively discussion on Hacker News.
-
Anthropic to Keep Claude Fable 5 Available in Max and Team Subscriptions
Anthropic has abandoned plans to remove the Claude Fable 5 model from its subscription tiers and will keep it available to Max and Team Premium users. The decision is tied to competition from GPT-5.6 Sol and Kimi 3.
-
Kaiser Permanente Nurses Say AI and Workplace Surveillance Are Making Patient Care Worse
Medical staff at Kaiser Permanente, a major U.S. clinic network, complain that the rollout of AI tools and employee-surveillance systems is hurting the quality of their work and of patient care. The story drew wide discussion on Hacker News.
-
Databricks Reaches a $188 Billion Valuation on the Wave of AI Interest
Databricks, once known as a big-data platform, has announced a new valuation of $188 billion, cementing its status as one of the leading beneficiaries of the artificial intelligence boom. The company also published research on the savings that come from using open AI models to write code.
-
Apple and Google Ordered to Remove “Nudify” AI Services from Their App Stores
San Francisco's city attorney has sent Apple and Google letters demanding that they pull “nudify” apps from the App Store and Google Play. According to him, both companies have long known that the apps break state law.
-
Apple's Trade Secret Suit Against OpenAI Could Complicate Its IPO Preparations
Apple has filed a trade secret misappropriation suit against OpenAI, accusing the company of systemic violations involving its head of hardware. The suit comes at an awkward moment for OpenAI — amid talk of a possible IPO.
-
The European Commission Orders Google to Open Access to Android's Microphone and Camera to Competitors
The European Commission has ordered Google to grant third-party AI assistants the same capabilities on Android that Gemini already enjoys. The changes are due to appear in Android 18 and to take effect no later than August 1, 2027.
-
Golf for Birds: An Ironic Recipe for Saving Data-Center Water
Blogger Simon Willison has proposed a tongue-in-cheek way for tech giants to offset criticism over data-center water use — buy up golf courses and turn them into birdwatching parks.
- More