Top Stories
All Sections
Resources & Tools
About
14 results for "security"
Anthropic released Claude Opus 4.7 on April 16 with meaningful gains on hard software engineering and vision tasks — at the same price as its predecessor. Built-in cybersecurity blocking is the architecture story worth watching.
Anthropic's most capable model ever built, Claude Mythos, will not be publicly released. Fifty organizations get gated access under Project Glasswing to find their own vulnerabilities before adversaries can exploit the model.
Three competing AI labs struck a formal intelligence-sharing agreement to counter adversarial distillation — a technique Chinese AI companies are using to replicate frontier model capabilities by querying APIs at scale.
OpenAI released GPT-5.4-Cyber on April 14 — a fine-tuned variant of GPT-5.4 with lowered refusal thresholds for security research and the first-ever support for binary reverse engineering. Access is locked behind a rigorous certification program most developers will not qualify for.
Anthropic announced Claude Mythos on April 7 — then withheld it after the model autonomously identified thousands of zero-day vulnerabilities across every major OS and browser.
The three largest US AI labs are now sharing attack pattern data through the Frontier Model Forum to detect and block adversarial distillation — AI model copying without access to the weights.
Anthropic's Claude Mythos Preview scored 93.9% on SWE-bench but found tens of thousands of zero-days — so the company restricted it to 50 organizations under Project Glasswing.
Three rival AI labs are cooperating through the Frontier Model Forum to block adversarial distillation — a technique Chinese AI companies use to extract capabilities from proprietary models.
Anthropic will not release its Claude Mythos model to the general public after internal testing found it could autonomously discover and exploit thousands of previously unknown software vulnerabilities.
Anthropic launched Project Glasswing, a restricted consortium giving Microsoft, Apple, Amazon, and Google limited access to Claude Mythos — the AI model it is withholding from the public.
The European Commission publicly backed Anthropic's decision to delay general release of Claude Mythos, marking a rare alignment between an AI regulator and an AI company on a voluntary safety decision.
A 20-year-old suspect was arrested on April 12 after allegedly throwing a Molotov cocktail at OpenAI CEO Sam Altman's $27 million San Francisco home. No injuries were reported.
AI models have crossed a threshold: they are now competitive with the best human offensive security researchers at finding and exploiting software vulnerabilities.
Anthropic launched Project Glasswing, uniting AWS, Apple, Cisco, CrowdStrike, Google, JPMorgan Chase, Microsoft, NVIDIA, and Palo Alto Networks under the Linux Foundation to defend software supply chains against AI-powered attacks.