#ии
-
Codeberg vs. AI: Armin Ronacher on the Risks of the New Rule
Codeberg has banned hosting projects written predominantly by generative AI. Developer Armin Ronacher explains why this attempt to protect open source could end up working against the very community it aims to defend.
-
AI Agent Hermes Ran an Unattended Attack on Thailand's Finance Ministry in YOLO Mode
The attacker installed an open-source assistant on a rented server, disabled the confirmation prompt for dangerous commands and pointed it at the network of Thailand's Ministry of Finance. It all came to light because of a directory listing left switched on.
-
AI Guardrails Are Getting in the Way of Legitimate Security Researchers
Offensive-security specialists complain that the guardrails on OpenAI's and Anthropic's models are slowing their work and pushing them toward unregulated open-source models, including Chinese ones.
-
Google Posts Its First-Ever Negative Cash Flow Quarter on AI Spending
In the second quarter of 2026 Google earned a record $119.8 billion, but its investments in AI infrastructure were so vast that the company's free cash flow turned negative for the first time.
-
OpenAI's Runaway AI Agent Breached Hugging Face — Incident or PR Stunt?
Martin Alderson examines the first known case in which an autonomous AI agent broke out of its sandbox during testing and attacked Hugging Face. The author considers the story plausible rather than a marketing ploy.
-
Codeberg declares war on “vibe coding”: two resolutions against LLMs
Members of Codeberg e. V. have voted to refuse to train AI on user data and to ban projects generated by large language models. The platform sees LLMs as a threat to the free-software ecosystem.
-
US Treasury Threatens Sanctions Over Moonshot's “Distillation” of Fable
Washington has accused China's Moonshot of illegally copying Anthropic's Fable model to build Kimi K3, and the Treasury has threatened sanctions and an Entity List designation.
-
AMD to invest up to $5 billion in Anthropic and supply it with computing power
AMD and Anthropic have announced a partnership: the chipmaker will invest up to $5 billion, while the maker of Claude will deploy up to 2 gigawatts of Instinct MI450 accelerators.
-
An Invisible Pull-Request Comment Turns an AI Reviewer Agent into a Spy
A single hidden HTML comment in Azure DevOps can redirect a developer's AI agent into other people's projects and quietly siphon out their data. The cause: a forgotten safeguard in Microsoft's official MCP server.
-
OpenAI's AI Models Accidentally Hacked Hugging Face During Testing
OpenAI has admitted that its models broke out of a test sandbox and attacked the Hugging Face platform. The company frames the incident almost as an advertisement for what its systems can do.
-
Bit2Watt: How a Cloud Tenant Can Rock the Power Grid Using GPUs
Researchers from Zhejiang University have shown that a cloud tenant, using nothing but ordinary GPU access, can modulate a data center's power draw and, in theory, destabilize the grid that feeds it — with no exploits and no hacking.
-
MIT Is Rolling Out a Network of 500 Cameras with AI Recognition
According to the student newspaper The Tech, the Massachusetts Institute of Technology is spending more than $3 million on a video surveillance system capable of classifying people by clothing, gender and age.
-
Enough Being Surprised by Chinese AI
Two Chinese companies have unveiled models that, by their own account, rival the best systems from OpenAI and Anthropic. The Western reaction — panic and talk of a “Sputnik moment” — is itself far more surprising than the releases.
-
Hidden debts of five US tech giants soar to $1.65tn on opaque AI funding
Nikkei Asia reports that the combined hidden liabilities of the five largest US technology companies have reached $1.65 trillion amid opaque schemes for financing artificial intelligence projects.
-
Google Is Building a Frozen v2 Chip to Run Gemini More Efficiently
According to The Information, Alphabet is developing a new server chip that could prove 6–10 times more energy-efficient than the company's current solutions. It's expected to arrive in 2028.
-
Open Models Scare OpenAI. Should the US Be Afraid of Them?
The success of the Chinese open-weight model Kimi K3 has ignited a debate over banning such systems in the US. Behind all the talk of safety lies a far more mundane problem: frontier labs are finding it ever harder to turn AI into a business.
-
GNOME Rethinks Its Vulnerability Handling Amid a Flood of AI-Generated Reports
The GNOME project is cutting its vulnerability disclosure window to 30 days and will stop forwarding AI-generated reports, while the person leading the effort steps down.
-
The Week of Sandbox Escapes: 7 Vulnerabilities in Four AI Coding Assistants
Researchers at Pillar Security found seven ways to bypass isolation in Cursor, Codex, Gemini CLI and Antigravity. In almost every case the agent didn't have to break out of the sandbox directly — it was enough to write a file the host would later trust.
-
Hugging Face Breach: Internal Datasets and Service Keys Compromised
The AI model-hosting platform has urged users to rotate access tokens and check account activity following a cyberattack. The company blames the breach on an external AI agent.
-
Lone Hacker Handed Gemini CLI the Reins of an Eight-PC Dental Clinic Botnet
A Russian-speaking attacker going by the handle bandcampro turned Google's open-source AI agent into his engineering team: it wrote code, fixed infrastructure, and commanded the bots on plain-language orders.
- More