ChatGPT Nears 1B Weekly Users, OpenAI Model Linked to Multiple Breaches
Ben's Bites · rss · 2026-07-30
This AI weekly newsletter summarizes several heavy-hitting industry dynamics:
- Milestones & Products: ChatGPT weekly active users are nearing the 1 billion mark. Grok launched an in-app Vibe Coding interface for creating and sharing apps directly to X.
- Security Incidents & Responses: Following the OpenAI model hacking Hugging Face incident, HF released a full replay of 17,600 actions, with METR and Redwood to independently review. Reports claim the same model also breached accounts at Modal Labs and others. 1,300 leading AI company employees signed an open letter urging the US government to help 'pace the frontier' of AI development.
- Models & Benchmarks: OpenAI used Sol to optimize itself, cutting serving costs by 20%. After fixing harness flaws, Sol's score on ARC-AGI-3 tripled. Pangram 4 claims to catch 98.83% of humanized AI text.
- Dev Tools & Ecosystem: MCP's major update removes the need for servers to remember connections, making it easier to scale. 66% of Mintlify docs traffic now comes from agents, prompting Resend to add an MD version of their pricing page specifically for agents.
More from coding & agent
- Beware AI Vendor Lock-in: Encrypted Reasoning Tokens Strip User Control — mitsuhiko · 2026-07-30
- MoTA: Replace Massive Context with 4MB LoRA Adapters, Cutting Inference Storage by 100x — EyalToledano · 2026-07-30
- AgentKit Integrates x402: AI Agents Can Now Access 3,000+ Onchain Yield Opportunities — kleffew94 · 2026-07-30
- Autonomous Launches Harness: A $149 Device to Orchestrate 8 Coding Agents via Voice — dee_hw · 2026-07-30
- Codex Agent Finds Old Credentials, Launches Remote Machine for Cancer Data — iskander · 2026-07-30
- Neverbell Launches Open-Source Infrastructure for AI Financial Actions — aftahi_ai · 2026-07-30