Lead
Meet GPT-Red: an LLM super-hacker OpenAI built to make its models safer
OpenAI has built an LLM super-hacker called GPT-Red that it uses as a sparring partner to help its other models boost their defenses against…
Daily brief
Wednesday · July 15 · 2026
12 stories · ~5 min
What matters in AI today. Curated for researchers, engineers, and founders.
Lead
OpenAI has built an LLM super-hacker called GPT-Red that it uses as a sparring partner to help its other models boost their defenses against…
Top stories
An AI industry professional tested ChatGPT Work and Claude Cowork on organizing PDFs, finding that while Work excelled in identifying and removing…
A new benchmark, Real World VoiceEQ, evaluates voice AI’s naturalness, reliability, and emotional understanding in real conversations, highlighting gaps…
Essay
All essays →“Founded by OpenAI veterans, Anthropic has developed Claude, a safer AI model using Constitutional AI, targeting regulated industries and enterprise….”
Read essay →Deep read
Original analysis — not aggregated from RSS.
Original analysis
All essaysIndustry
Founded by OpenAI veterans, Anthropic has developed Claude, a safer AI model using Constitutional AI, targeting regulated industries and enterprise…
Industry
OpenAI has transformed from a non-profit research lab into a leading platform company in the Generative AI era, leveraging a 'capped-profit' model and a…
Industry
Mistral AI, founded by former DeepMind and Meta researchers, is Europe's leading generative AI, championing 'open-weight' models that prioritize…
Newsletter
Five AI stories. One essay pick. Three minutes. Free forever.
Gidi Littwin, co-inventor of Apple’s FaceID, has launched Hemispheric, a startup using AI to decode brain activity for diagnosing cognitive disorders…
OpenAI Staffers Are Funding a Rival Super PAC to Take on Their Boss