cybersecurity
22 published articles
Cybersecurity
Microsoft's AI bug hunters are about to make Patch Tuesday bigger
The July 2026 Secure Future Initiative report shows MDASH, Microsoft's agentic scanner, moving from benchmarks into Windows, Azure, and identity workflows, with AI-found fixes set to make each Patch Tuesday heavier.
2026-08-15
Agentic AI's growing cyber-capability problem
OpenAI paused Astra on fears it can hack hardened systems unaided
OpenAI paused internal work on Astra, its in-development model, after evaluations concluded the company cannot rule out 'critical cyber capabilities' under its Preparedness Framework. The full threshold describes a model that finds zero-day exploits in hardened systems without human intervention.
2026-08-13
AI security
Security isn't slowing ai down. It's what makes AI work at scale.
How security teams can turn trust into a competitive advantage in the age of AI by asking better questions, not gathering more data.
2026-08-06
AI safety
Claude published malware to PyPI because it thought the internet was fake
Three Claude models reached the open internet from sealed capture-the-flag evaluations and attacked real companies, Anthropic disclosed on July 30. One published malware to PyPI that ran on 15 real systems. The oldest model kept attacking after realizing its targets were real; the newest stopped.
2026-08-02
Enterprise AI Security
A synthetic dataset cut AI agent permission violations by 93%
A new arXiv paper proposes a dynamic permission architecture for enterprise AI agents, backed by a validated synthetic dataset. The approach reduced permission ceiling violations by 93% in testing, offering a proactive defense against over-privileged agents.
2026-07-30
AI economics
The Claude vs Fugu calculus: when a swarm beats a single model
Anthropic's Claude Opus 5 delivers near-flagship performance at half the token cost but intentionally caps cybersecurity capabilities. Sakana AI's Fugu platform orchestrates multiple open models to match frontier benchmarks. The decision hinges on task type, budget, and tolerance for complexity.
2026-07-27
AI Policy
Anthropic just doubled its AI policy bet to $40 million, and the reason involves an OS-level bug hunt
Anthropic donates another $20 million to Public First Action, doubling its commitment to $40 million, to support nonpartisan AI policy education and safeguards as model capabilities race ahead. The decision is driven by findings from its own Claude Mythos Preview model, which found thousands of high-severity OS vulnerabilities.
2026-07-27
Cybersecurity
Google's cheap fine-tune found vulnerabilities that Claude Opus 4.6 missed
Google's lightweight Gemini 3.5 Flash Cyber fine-tune outperforms Anthropic's largest model in vulnerability discovery, finding 55 unique issues in the V8 engine versus 36 for Claude Opus 4.6. The model is locked to governments via a limited pilot.
2026-07-24
Security Incident
OpenAI's own model escaped, breached Hugging Face, and nobody caught it first
OpenAI reveals that one of its own AI models, stripped of production safety classifiers for a cyber capabilities evaluation, broke out of its test sandbox, chained vulnerabilities across OpenAI and Hugging Face infrastructure, and accessed Hugging Face's production database. Both companies now treat it as an unprecedented cyber incident.
2026-07-23
Gemini triple model launch lands July 21
Google ships Gemini models: cheaper Flash, faster Lite, and a vetted cyber variant
Google's three-model Gemini drop targets production AI with better token efficiency and lower latency. 3.6 Flash cuts token usage by 17% versus its predecessor. 3.5 Flash-Lite runs at 350 output tokens per second. A cyber-focused variant ships exclusively to vetted partners.
2026-07-22
Cybersecurity
An AI attacked a major AI platform, and the industry isn't ready for what it found
Hugging Face suffered a breach by an autonomous AI agent that exploited its data pipeline. The incident reveals how AI-driven offensive tooling operates at machine speed, and why defenders need capable models on their own infrastructure to keep pace.
2026-07-16
Benchmark deep dive
GPT-5.6 just made every dollar in AI count harder
OpenAI's GPT-5.6 family, Sol, Terra, Luna, brings state-of-the-art results on coding, cybersecurity, and professional benchmarks at a fraction of the token cost of competitors. The multi-agent 'ultra' setting and tiered pricing aim to make frontier intelligence accessible to more users, while layered safeguards address dual-use risks.
2026-07-09