Former MIT scientist Joscha Bach funded by Epstein made racist claims in private (www.bostonglobe.com)

Newly released documents from the U.S. House Oversight Committee show that Joscha Bach, a former MIT Media Lab researcher whose position was partly funded by Jeffrey Epstein, exchanged emails with Epstein in 2016 containing racist and sexist generalizations — claiming, for example, biologically rooted cognitive gaps between Black and white children and attributing women’s lower representation in math and computer science to intrinsic disinterest in abstract systems. The correspondence, which also veered into musings about fascism and Epstein’s “political incorrectness,” deepens scrutiny of Epstein-linked donations: Epstein subsidized Bach’s Media Lab appointment (about $300,000 in gifts), introduced him to other academics, and facilitated access to Harvard’s Program for Evolutionary Dynamics. Bach says his views on race have since changed; institutions including MIT and Harvard have faced reviews and taken remedial steps, and a startup that had engaged Bach severed ties after the emails surfaced. For the AI/ML community the episode underscores two interlocking risks: the reputational and ethical consequences of opaque funding relationships, and how researchers’ unexamined biases can undermine trust in AI research and its social impact. Technically, it highlights why rigorous attention to bias, diverse datasets, and ethical governance matters not just for models but for who sets research agendas. The revelations reinforce calls for stronger gift‑acceptance policies, transparency about donor influence, and community norms that safeguard scientific rigor and inclusivity in AI research and hiring.