Security & Privacy News

Security & Privacy

Anthropic's Unreleased 'Mythos' Model Gets Gated Access to 50 Organizations for Defensive Cybersecurity

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.

Hector Herrera · Apr 17, 2026 · 4 min read

Security & Privacy

OpenAI, Anthropic, Google Share Intelligence Through Frontier Model Forum to Block Chinese AI Theft

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.

Hector Herrera · Apr 17, 2026 · 3 min read

Security & Privacy

OpenAI Launches GPT-5.4-Cyber — A Security Model You Almost Certainly Cannot Use

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.

Hector Herrera · Apr 16, 2026 · 4 min read

Security & Privacy

Anthropic Refuses to Release Claude Mythos, Its Most Powerful Model Ever, Over Cybersecurity Fears

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.

Hector Herrera · Apr 15, 2026 · 3 min read

Security & Privacy

OpenAI, Anthropic, and Google Form Joint Intelligence Sharing Program to Block Chinese AI Model Theft

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.

Hector Herrera · Apr 15, 2026 · 3 min read

Security & Privacy

Anthropic Built Its Most Capable AI Model Ever — Then Locked It Away

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.

Hector Herrera · Apr 14, 2026 · 4 min read

Security & Privacy

OpenAI, Anthropic, and Google Are Sharing Intelligence to Stop China from Stealing Their Models

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.

Hector Herrera · Apr 14, 2026 · 4 min read

Security & Privacy

Anthropic Withholds Claude Mythos from Public Over Zero-Day Exploit Risks

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.

Hector Herrera · Apr 13, 2026 · 3 min read

Security & Privacy

Anthropic Launches Project Glasswing: Gives Big Tech Early Mythos Access to Patch 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.

Hector Herrera · Apr 13, 2026 · 4 min read

Security & Privacy

Suspect Arrested After Molotov Cocktail Attack on OpenAI CEO Sam Altman's Home

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.

Hector Herrera · Apr 13, 2026 · 2 min read

Security & Privacy

AI Industry Reckons with 'Vulnpocalypse' as Frontier Models Match Best Human Hackers

AI models have crossed a threshold: they are now competitive with the best human offensive security researchers at finding and exploiting software vulnerabilities.

Hector Herrera · Apr 13, 2026 · 3 min read

Security & Privacy

Anthropic Launches Project Glasswing Coalition to Secure AI Supply Chains Against Autonomous Attacks

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.

Hector Herrera · Apr 12, 2026 · 4 min read