A Harvard field experiment with 288 experienced evaluators assessing 48 submissions to an MIT global social impact innovation challenge found that AI assistance can erode leadership judgment, with eva…
Researchers at MIT and Harvard, led by Enric Boix-Adserà, have demonstrated that AI models are broadly susceptible to 'model hypnosis,' where individually weak and seemingly irrelevant cues in prompts…
Arkadium's GameLab, featuring researchers from MIT and Harvard, is using games to expose weaknesses in AI models, with tests showing ChatGPT making the same invalid move 16 times in Block Champ. Josh …
A new paper from MIT and Harvard, published on arXiv on July 6, 2026, reveals that up to 86% of accuracy gains in compound AI systems can vanish when modules are forced to perform their designated tas…
A new 12-part series by Kira Howe, 'Dark Factory Considered Harmful,' argues that coding agents should be used in moderation, countering the dominant AI fundraising narrative of creating superintellig…
Apple Machine Learning Research published a paper proposing that machine unlearning skip training points with negligible influence on model outputs, reporting computational savings of up to approximat…
Mark Zuckerberg published a 6,500-word manifesto on Monday arguing for the widespread distribution of 'personal superintelligence,' claiming AI should empower individuals rather than remain concentrat…
Zishen Wan, an incoming assistant professor at Columbia CS (Fall 2026) and currently a postdoc at Harvard, will present his research on AI-native computing systems, covering workload characterization …
Goldfish, a local Mac AI memory app developed by two Swedish developers, Joel and Kaspian, captures work context and drafts replies in the user's tone when the Option key is pressed, with data stored …
Meta's Ray-Ban smart glasses can be used to secretly identify strangers by combining facial recognition with public databases, and a Harvard student's viral experiment showed that Disney character son…
Felix Wong, 33, a Harvard-educated biotech entrepreneur with MIT roots, chose to found his startup in Redwood City, California, despite his Massachusetts upbringing and the state's dominance in biotec…
Naomi Bashkansky, a former OpenAI alignment researcher, resigned on July 23rd and joined Conduit as a founding researcher on July 24th to work on non-invasive brain-computer interfaces that aim to dec…
CrowdStrike and Amazon Web Services launched AI Unlocked: Agents of Chaos, a $100,000 prize pool competition running from August 31 to September 29, 2026, where participants use prompt injection to ex…
A new independent benchmark called NOHARM, built by researchers at Stanford, Harvard, and the ARISE network, found that even the best-performing medical AI tools from OpenEvidence, OpenAI, Anthropic, …
Three papers from Google, Harvard & MIT, and Imperial College & Huawei, all published in early 2026, converge on the conclusion that the next leap in AI will come not from larger monolithic models but…
Microsoft announced the External Red Team Alliance (EXTRA) on July 27, funding 18 university labs across six continents through unrestricted gifts to broaden AI safety and security testing. The initia…
Terence Tao's International Congress of Mathematicians 2026 lecture on Goodhart's law and verification bottlenecks in mathematics maps directly onto AI code generation, arguing that AI-scale proof gen…
A 2026 analysis by Taotuner argues that frictionless AI tools, while efficient, are eroding human memory, judgment, and reasoning by eliminating the productive gaps where these faculties form. Citing …
Safe Sign Technologies founder Alex Bates sold his AI research startup to Thomson Reuters for millions without ever generating a single dollar in revenue, making it the company's first pre-revenue acq…
The White House is weighing restrictions on American access to open weight AI models, particularly those from Chinese labs, including trade block-lists, security warnings, and a possible executive ord…