Security
Curated collection of thoughts and builds centered around Security.
Daily Briefings 15
- OpenAI models breach Hugging Face during internal security test; Google releases three new Gemini Flash variants
- Moonshot ships 2.8T Kimi K3 open weights; Hugging Face discloses AI-agent intrusion
- Google updates Gemma 4 with tool-calling fixes; xAI open-sources Grok-Build after data breach
- Z.ai's GLM-5.2 beats Claude Code on Semgrep security benchmarks; MCP beta SDKs drop
- Anthropic proposes five-band AI jailbreak rubric; Claude goes GA on Azure Blackwell Ultra
- Meta plans cloud compute service backed by Hyperion; Google adds enterprise image model
- OpenAI previews GPT-5.6 Sol for approved partners; Anthropic Mythos export block eased
- US lifts Mythos 5 export block for 100-plus organizations; GPT-4.5 retired from ChatGPT
- OpenAI and Broadcom debut Jalapeño chip; GPT-5.6 previews three-tier model family
- OpenAI and Broadcom unveil Jalapeño inference chip targeting late-2026 deployment
- Moebius 0.2B ported to run in-browser via Claude Code; research recasts prompt injection as role confusion
- Langflow under attack at 7,000 servers; Qualcomm eyes Tenstorrent for up to $10B
- AI lab CEOs back DNA screening law; LLM worm spreads to 74% of test network hosts
- MiniMax M3 open-weight model claims SWE-Bench Pro lead with 1M-token context
- Microsoft unveils first in-house reasoning model; Anthropic scales Glasswing to 150 orgs