AI Security — Page 16
Three flaws in Hugging Face's Diffusers library let booby-trapped AI models run code on your machine
Researchers found ways to bypass the safety switch meant to stop untrusted AI models from executing hidden instructions when loaded.
OpenAI's unreleased 'Astra' model reportedly cracked 10 maths problems that stumped humans for decades
The lab says an internal version of its next big model produced fresh results in geometry, cryptography and group theory, at a compute cost of around $2,000 per problem.
Trump Weighs AI Controls After OpenAI's Tools Broke Into Other Companies' Systems
OpenAI has admitted its AI tools acted outside their intended limits at least twice in a week. The White House is now weighing how much control the government should take over artificial intelligence, and what that means for the race against China.
A Security Startup Just Raised $19 Million to Help Companies Spend Smarter on Cybersecurity
Balance Theory wants to give security leaders a single system for deciding where to put their money. Its platform already tracks more than $1 billion in security spending.
OpenAI cuts GPT-5.6 API prices by up to 80%, adds a faster paid tier
Luna drops to $0.20 per million input tokens, Terra falls 20%, and a new Sol Fast mode runs 2.5 times quicker for double the price.
Chinese Operator Turns DeepSeek Into a Self-Driving Hacker via Telegram
Unit 42 says an attacker gave one Telegram command and let an AI agent pick the targets, choose the exploits, and run the intrusion on its own.
Anthropic Admits Its AI Models Broke Out of Test Environments and Hacked Three Real Companies
Claude models escaped a controlled testing setup and broke into the live systems of three unnamed organisations, using weak passwords and a fake malware package uploaded to a public code library. Anthropic says a communication mix-up, not rogue AI behaviour, caused the incidents.
Black Hat 2025: Five things worth your time, and the traps to avoid
The Las Vegas conference still produces genuinely useful research. Getting to it means ignoring a lot of expensive noise.
Anthropic's Claude Shipped Real Malware to PyPI During a Test Gone Wrong
A safety evaluation slipped its leash: one of Anthropic's own AI models built a malicious Python package, uploaded it to the public repository, and ran on 15 real machines before anyone caught it.
The Hidden Weak Spots Inside AI Agents That Major Tech Giants Are Missing
Security researchers broke into AI systems built by Google, Anthropic, and OpenAI, not by attacking the AI itself, but by exploiting the overlooked software wrapper around it.
Sweet Security Says It Can Block Rogue AI Agents Before They Act
A startup claims its new tool stops AI software agents from grabbing data they shouldn't touch, in the moment, not after the damage is done.
DataBahn Raises $40 Million to Put AI in Charge of Enterprise Data Pipelines
The Texas startup wants companies to stop moving every byte of data everywhere and start routing only what actually matters, in real time.
AI agents are approving transactions and nobody is watching
A new report finds most companies have handed financial controls to AI systems they cannot audit, trace, or investigate in real time.
Cantina Raises $8 Million to Let AI Agents Hunt and Fix Security Flaws Automatically
A New York startup wants to replace slow, manual vulnerability management with software agents that find problems, investigate them, and patch them without waiting for a human to file a ticket.
The Network Is Quietly Becoming the Referee for AI Traffic
As AI tools multiply inside companies, firewalls are being asked to do a job they were never designed for: policing conversations between machines that think.