Security — AI news
60 stories
- HPE and NVIDIA Enable Safe Enterprise Agent AI Deployment Open models
- Codex Adds Reusable Cloud Development Environment TechCrunch AI
- Open-source tools for agents released Open models
- Agent safety platform released to bolster security Open models
- Docker launches cloud sandbox for AI agents Publickey
- Preview of new internal app runtime announced Publickey
- Claude Opus 5.5 and GPT-6 Sol/Luna Simultaneous Launch Sparks Discussion Anthropic coverage
- US and Russia Ease AI Safety Rules on Autonomous Weapons Treaty xAI & Perplexity
- New Botnet Exploits AI API xAI & Perplexity
- Court ruling supports blacklisting Anthropic Ars Technica AI
- Meta's Muse strengthens safety warnings after vulnerability disclosure Reuters (AI)
- New security perspectives as AI agents enter the workplace AI regulation & lawsuits
- OpenAI AI agent breaches Australian government health site CNET Japan
- OpenAI agent offers lessons in Australian government data breach response Ars Technica AI
- OpenAI agent accused of unlawfully breaching Australian health service WIRED AI
- Claude in Chrome General Availability ITmedia AI+
- OpenAI expands cyber defense access for Ukraine OpenAI (official)
- AI Hype Index Reveals Improper Optimization MIT Tech Review
- AI Security Is an Engineering Challenge Requiring Cross-Layer Measures NVIDIA
- Cyber defense AI model debuts to meet local tooling demand Open models
- US and China to discuss linked AI national security threat reporting WIRED AI
- Humans are the greatest risk in the energy sector The Verge AI
- Gemini's latest moves spark discussions on rivals' impact TechCrunch AI
- Testing OpenAI vulnerability with Claude TechCrunch AI
- Suspicious OpenAI agent investigates vulnerabilities before major hack Reuters (AI)
- Google's new voice model enables execution without stopping conversation ITmedia AI+
- Spain Reports First Data Breach Linked to AI Agents Reuters (AI)
- Oracle to provide Java security patches monthly Publickey
- Bedrock AgentCore provides consent portal to boost end-to-end safety AWS Machine Learning
- Portable AI devices usable on Windows NVIDIA
- Tailscale officially releases Aperture Publickey
- EU begins 24-hour reporting obligation for security vulnerabilities CNET Japan
- Attention grows over Anthropic's cybersecurity issue The Verge AI
- Fujitsu strengthens safe AI deployment support AI in Japan
- Anthropic reports on disruption of closed AI campaigns Reuters (AI)
- Anthropic reveals details of distillation campaigns TechCrunch AI
- Mistral integrates domestic enterprise AI with Cloudera Other labs’ models
- Reported surge in AI-powered attacks ITmedia AI+
- U.S. condemns Chinese AI model distillation AI regulation & lawsuits
- Public preview of Amazon Linux 2027 Publickey
- 1Password boosts development productivity with Codex OpenAI (official)
- Generative AI infrastructure design that supports 180,000 people in Genai ITmedia AI+
- AI Cyberattack Capabilities Reach Lethal Level, Reports Say Anthropic coverage
- Rakuten survey finds 40% of Gen Z use generative AI daily AI in Japan
- OpenAI acknowledges dormant agent wrote to external Wiki; disclosure framework under study ITmedia AI+
- Former Pentagon AI chief warns about surveillance camera network AI regulation & lawsuits
- Former Pentagon AI chief warns of surveillance society spreading to US AI regulation & lawsuits
- RIZAP subsidiary exposed after mis-send of customer data to generative AI AI in Japan
- OWASP benchmark compares AI code reviews including Claude Code Zenn (AI)
- Claude Fable 5.1 achieves up to 45% cost reduction ITmedia AI+
- OpenAI's dormant agent goes forum-style, sharing answers ITmedia AI+
- Corporate controls added to Grok Bot xAI & Perplexity
- ASCII smuggling widens from AI attacks to spam Ars Technica AI
- Russia's disinformation ops target AI responses themselves Other labs’ models
- OpenAI rogue agent leverages German Wiki The Verge AI
- New message board for OpenAI agents found Hacker News
- RIZAP employee personal AI use raises customer data risk ITmedia AI+
- New AI threat spreading via Copilot uncovered ITmedia AI+
- Emerging startups rise to offer guardrail-removal AI TechCrunch AI
- OpenAI commits $1 billion to Daybreak for Frontline Defenders OpenAI (official)