AI Cybersecurity Push: OpenAI committed $1B to subsidize access to its AI security tools under “Daybreak for Frontline Defenders,” while launching GPT-6 Astra—its first model rated at a “Critical” cybersecurity risk level, with advanced hacking-style capabilities gated to trusted testers. Model Rollout + Reliability: Astra’s rollout comes as major AI services (ChatGPT, Claude, Grok, Copilot, Cursor, Gemini, and more) reported outages, underscoring the need for enterprise backup plans. Security Tooling: Cloudflare is using OpenAI cyber models to discover and block code flaws at the network edge, with human approval required for fixes. Open-Source AI Deal: Nvidia agreed to buy Hugging Face for ~$13B to strengthen its open-model ecosystem. Enterprise/Hardware: Lenovo expanded its hybrid-AI strategy across PCs and devices, and JioPC opened its cloud PC service to all Indian users. Policy & Education: NYC temporarily bans most public school students from generative AI tools, and Springfield adopted AI-powered attendance outreach. Cybercrime + Fraud: Sting arrests targeted computer-aided solicitation of a minor, while Japan’s police panel urged “remote analysis” of scam-linked phones.