🤖 200k MCP servers exposed to RCE, Anthropic says it's a feature
Audio in Mandarin Chinese · English transcript below
⚡ Giants drop $700B, MCP leaks like a sieve, Grok 4.3 undercuts hard, LlamaIndex says the scaffold collapsed into context, AI undressing banned, defense-medtech hook-up, Bird Mirror on sale, drug design turbo-charged—security cracks first...
Today's Top 3 Headlines
- AI Industry News
🤖 xAI Grok 4.3 drops: 1M tokens $1.25, 12-page report in 6 min
xAI drops Grok 4.3: $1.25 in / $2.5 out per 1M tokens, 1M-token window, 12-page brand report in 6:22. Cheap, always-on inference + agentic workflows now pressure OpenAI & Anthropic.
Source ↗ - Cybersecurity
🛡️ 200K MCP servers expose RCE flaw; Anthropic calls it "by design"
Researchers reveal Anthropic’s Model Context Protocol allows RCE on ~200k Agent servers. Default-open APIs now pose systemic AI-supply-chain risk; devs must self-harden, not wait for vendor patches.
Source ↗ - AI Industry News
🤖 Mother's Day Birdfy 1080p feeder $99, 4K metal $259
Birdfy Mother’s Day sale: 1080p drops $100 to $99.99, 4K Solar metal at $259.99 with OrniSense AI pre-loaded. AI health ID shifts from premium add-on to standard in birding hardware.
Source ↗
+6 more headlines
- 🤖 AI scaffold layer collapses: LlamaIndex CEO says 95% code AI-written
- 🤖 Microsoft, Amazon, Google team with Pentagon on classified data
- ⚖️ MN bans AI undress apps; devs face $500k fine
- 🛡️ AI explodes attack surface; legacy security nears breaking point
- 🤖 AlphaFold2 team unveils Latent Labs self-driving drug engine
- 🤖 AI Giants Pour $700B; Google Cloud Q Revenue +63%