Coefficient Giving donates $1B to GiveWell, reshaping EA with AI wealth

Coefficient Giving (cG) announced a $1 billion gift to GiveWell on July 23rd, 2026, increasing a prior $175M commitment. This single donation is roughly the size of everything GiveWell expects to grant this year. Both organizations have lowered their funding thresholds—GiveWell from 8x to 6x its benchmark, and cG from 2,000x to 1,000x—to reflect the new influx. The piece notes that "much of the future funding depends on the valuations of extremely volatile assets," a clear reference to AI wealth. cG explicitly states that a belief in AI's potential to transform global health influences their decision. This marks the first major example of the "Funding Anthropalypse" reshaping effective altruism, where AI-driven fortunes now directly fund global-health interventions like bednets and vitamin distribution.

For smaller donors (giving thousands annually), little changes—GiveWell's 6x benchmark remains excellent marginal cost-effectiveness, and GiveWell will signal if money should go elsewhere. For donors giving millions, the calculus shifts: the highest-impact move may be building "absorption capacity" in other areas rather than adding to GiveWell's margin. cG acknowledges that redirecting marginal donations would be "perfectly rational." However, this does nothing to address power concentration—cG and GiveWell are the two largest EA players, now both exposed to the same volatile AI wealth source. The grant itself is outstanding at the object level, but the fragile architecture of the space remains unaddressed.