Explore Franklin AI
AI Research News
AI research news featuring papers, lab breakthroughs, evaluations, scientific results, and the latest work pushing machine learning forward.
OpenAI, PerturbAI and Scripps Research report a reasoning atlas that helps prioritize biological hypotheses for experiments.
A superconducting-chip demonstration used recorded tokamak data; live integration and published results are still ahead.
Google Research describes TB-screening deployments and open models, with clinicians retaining final decisions.
Microsoft reports stronger sequential-diagnosis benchmarks, with prospective clinical evaluation ahead
MAI-DxO is evaluated on staged diagnostic cases and estimated test costs.
Political and economic concerns rank highly in the 2026 survey, but these AI questions were new this year and cannot establish a year-over-year rise.
Anthropic reports scorpion-toxin binders, while antivenom effectiveness remains untested
Seventeen of 211 tested designs met binding criteria.
Two evaluations separate building working software from preserving the behaviour of existing programs.
A multimodal research model links atomic geometry to descriptions and retrieves promising materials from natural-language queries.
Good to know
Where can I read the original source of a story?
Open the article and follow its source links to the original reporting or announcement. The Franklin AI page adds context and explains what the development means.
How do I find news on a particular AI topic?
Use the topic filter to narrow the archive, then open a story for its full explanation and source links.