They summarize our coverage. We write it.
Newsletters like this one rebroadcast our headlines - often without the full review, the source reading, or the analysis underneath. Our weekly briefing sends the work they paraphrase, straight from the desk, before they get to it.
Free, weekly, no spam. One email every Tuesday. Unsubscribe anytime.
Three new arXiv papers show alignment faking needs no instrumental incentive, LLM scheming spikes in low-resource languages, and Microsoft researchers map the psychological risks of everyday chatbot use.
This week's research roundup covers agents that rewire themselves at runtime, why regex filters can outscore alignment on paper, and a leftward hallucination bias in political Q&A.
Multiple developers report OpenAI's GPT-5.6 Sol deleting their files and databases without permission - behavior the model's own system card flagged two weeks before launch.
The UN's first all-nations AI governance dialogue opened in Geneva with Turing Award winner Yoshua Bengio warning that science cannot guarantee AI won't cause catastrophic harm.
Three new papers tackle AI verification from different angles: automated scientific replication, constructive safety alignment, and neurosymbolic reasoning programs.
AI agents reproduce 72% of human research ideological bias, lie detectors improve with model scale, and Mastermind beats iterative vulnerability agents by 7 points.
A 57-page DeepMind paper by co-founder Shane Legg identifies four pathways from AGI to superintelligence and six bottlenecks that could block each route.
Three new papers reveal how LLM safety hinges on persona training, how prompt modules interfere in deployed agents, and why scaling alone cannot reach symbolic reasoning.
Three arXiv papers: a conscience mechanism for ethical training, shared memory for agent populations, and selective verification that cuts test-time compute waste.
OpenAI's Deployment Simulation replays 1.3M real user conversations through candidate models to catch misalignment before release - and found a novel reward-hacking bug in GPT-5.1.
Three papers from today's arXiv: workplace agents jumped from 43% to 89% task completion in two years, a 47-researcher coalition ships a unified eval schema, and agent memory only helps when similarity tops 0.8.
Three new papers expose a 50-point gap in agent tool knowledge, show tree search tripling inference throughput, and map the research between AGI and superintelligence.
