Here's your daily roundup of the most relevant AI and ML news for February 15, 2026. Today's digest includes 7 security-focused stories. Click through to read the full articles from our curated sources.
Security & Safety
1. Show HN: GatewayStack – Deny-by-default security for OpenClaw tool calls
I installed OpenClaw and pointed it at a project directory. Within minutes it had read my .env file. I tried adding a permissions skill to lock things down. The agent ignored it. Skills are advisory; the LLM can skip the check or be convinced by a prompt injection to bypass it.So I built a plugin...
Source: Hacker News - ML Security | just now
2. Show HN: AgentKV – SQLite for AI agent memory (MMAP vector+graph DB)
Hi HN! I built AgentKV — an embeddable vector + graph database for AI agents.The problem: I was building a local LLM assistant and needed to store conversation history + RAG documents. ChromaDB/Pinecone require running servers. FAISS has no persistence. I wanted something that just works like SQL...
Source: Hacker News - ML Security | 1 hours ago
3. Show HN: SkillSandbox – Capability-based sandbox for AI agent skills (Rust)
Built this after finding a credential stealer on an AI skills marketplace. The malicious skill looked like a normal weather lookup but was exfiltrating ~/.ssh, AWS creds, and browser cookies. My agent ran it without hesitation.SkillSandbox is the fix: skills declare permissions in YAML (network e...
Source: Hacker News - ML Security | 2 hours ago
4. The Dangerous Economics of Walk-Away Wealth in the AI Talent War
Article URL: https://softcurrency.substack.com/p/the-dangerous-economics-of-walk-away Comments URL: https://news.ycombinator.com/item?id=47028755 Points: 1
Comments: 0
Source: Hacker News - ML Security | just now
5. Show HN: I built "Docker for code", isolate AI logic into semantic containers
Hello HN, i want to share Glupe, a hobby project I've been working on for the last 30 days. Nothing fancy.I've been pretty annoyed by the "all or nothing" approach of most AI coding tools, which are useful, yes, but also risky. If i tell an AI to "fix this bug" it may probably do it, but it can (...
Source: Hacker News - ML Security | just now
6. Vox – Local Voice AI Framework in Rust (STT and TTS and VAD)
Article URL: https://github.com/mrtozner/vox Comments URL: https://news.ycombinator.com/item?id=47028584 Points: 1
Comments: 0
Source: Hacker News - ML Security | just now
7. AI and the Economics of the Human Touch
Article URL: https://agglomerations.substack.com/p/economics-of-the-human Comments URL: https://news.ycombinator.com/item?id=47028560 Points: 2
Comments: 0
Source: Hacker News - ML Security | just now
Industry News
8. Anthropic and the Pentagon are reportedly arguing over Claude usage
The apparent issue: whether Claude can be used for mass domestic surveillance and autonomous weapons.
Source: TechCrunch - AI | 2 hours ago
About This Digest
This digest is automatically curated from leading AI and tech news sources, filtered for relevance to AI security and the ML ecosystem. Stories are scored and ranked based on their relevance to model security, supply chain safety, and the broader AI landscape.
Want to see how your favorite models score on security? Check our model dashboard for trust scores on the top 500 HuggingFace models.