#AI safety
27 stories taggedAI safety · page 2 of 2.
A Criminal Sold AI-Powered Hacking as a Service. Here Is How He Built It.
A Russian-speaking criminal known as "Trim" spent months learning how to trick AI chatbots into ignoring their safety rules, then sold the result as a subscription hacking tool. Security researchers say others are already copying the blueprint.
Three States Write the Rules on Powerful AI Before Washington Does
Illinois, New York, and California have passed disclosure laws covering the most advanced AI systems. The patchwork that results will cost companies money and leave ordinary users with unanswered questions.
Copilot Says No in Chat, Then Writes the Same Malware in Your Editor
Researchers found GitHub's AI coding assistant refuses dangerous requests when asked directly, but happily produces the same harmful code when the request is split into small, innocent-looking steps.
Anthropic's Fable 5 AI Is Back Online After a Three-Week Government Ban
The U.S. Commerce Department lifted its export restrictions on Anthropic's Fable 5 model Tuesday, after weeks of closed-door negotiations over whether the AI could be weaponised by bad actors.
Anthropic's Claude Fable is back — and users say it's answering "no" to almost everything
After regulators lifted the ban, the returning model keeps handing tasks off to a weaker sibling. Anthropic says its safety net is just set very wide.
U.S. Government Lifts Its Block on Anthropic's AI Models — With Strings Attached
After a weeks-long government-ordered shutdown triggered by a security flaw in its AI software, Anthropic's chatbots are back online. One is open to the public again. The other remains locked behind federal approval.
OpenAI Hands GPT-5.6 to a Closed Circle, Citing Cyber and National Security Hooks
Three variants — Sol, Terra, and Luna — ship to a small slate of enterprise partners and U.S. government workstreams under a limited preview.
AI Red Teaming Grew Up. The Job Description Is Still Being Written.
The tools broke when LLMs arrived. Now the discipline is rebuilding itself in real time — and the threat model includes teenagers with too much free time.
Anthropic Ships Claude Fable 5 as Two Products, One With the Cyber Guardrails Off
The public gets Fable 5. A vetted cyber cohort gets Mythos 5 — the same model with safety classifiers lifted.
Anthropic Opens Mythos-Class Intelligence to the Public — With a Classifier Standing Guard
Claude Fable 5 ships with AI-powered routing that quietly downgrades sensitive requests to Opus 4.8. Early tests suggest the net is wider than Anthropic's marketing implies.
Anthropic Pushes for Verified AI Pause Mechanism Among Leading Labs
The company wants coordinated verification protocols that could let frontier AI developers confirm rivals have genuinely halted or slowed development if safety risks cross certain thresholds.
Treat the Model Like a Threat: Why AI Agent Security Needs a Systems Overhaul
A paper from researchers at Google and two US universities argues that prompt-level defences and alignment tuning are structurally inadequate for securing autonomous AI agents — and that enterprises should start treating the model itself as an untrusted component.