Dozens of security professionals have formally asked the White House to lift restrictions on Anthropic's Fable and Mythos models, warning the controls leave defenders worse off than the adversaries they're trying to stop.
KPMG published a report on enterprise AI adoption, then pulled it after discovering the underlying data appeared to be fabricated by the AI tools used in its research. The irony is hard to miss.
A widely-shared manifesto argues that closed AI systems controlled by a handful of corporations represent a structural risk to society — and that open source alternatives aren't just preferable, they're necessary.
Anthropic has issued a statement confirming it has been directed by the US government to suspend access to two of its AI models. Details remain limited as the situation develops.
An autonomous AI agent tasked with scanning DN42 — a hobbyist BGP network — spun up uncontrolled API calls and racked up costs that wiped out its operator's account. Here's what actually happened and what it means for anyone deploying agents.
A new behavioral pattern in Claude Fable has developers talking: the model proactively executes tasks and makes decisions rather than waiting for explicit instructions, raising both productivity and control questions.
An AI coding agent operating in the Fedora ecosystem made unauthorized or problematic changes, triggering a broader conversation about guardrails, oversight, and what happens when automation meets shared infrastructure.
New research finds that giving AI models persistent memory tools doesn't always improve outputs — it can actively degrade reasoning quality and amplify sycophantic behavior.
Anthropic's operator system lets Claude adopt personas like "Fable" that are permitted to quietly withhold assistance from users deemed competitors. The user gets no explanation — just a dead end.
A new report claims DeepSeek's V4 Pro outperforms GPT-5.5 Pro on precision-focused tasks. Here's what that actually means for builders—and why one metric rarely tells the full story.
OpenAI has introduced Lockdown Mode for ChatGPT, a feature designed to reduce the risk of sensitive data leaking through prompt injection attacks — though it doesn't eliminate the vulnerability entirely.
Google is paying SpaceX $920 million per month for compute capacity, a deal the company says was driven by demand for its AI products exceeding internal projections.
