organization
AI safety metaTechnical AI safety researchOther
| Foresight Institute | $150,000 | 73% | 1 |
| Anton Makiievskyi | $50,000 | 24% | 1 |
| Marius Hobbhahn | $2,000 | 1% | 1 |
| BlueDot Impact | $1,200 | 1% | 1 |
| Austin Chen | $1,000 | 0% | 1 |
| Date ↓ | Funder | Via | Amount | Cause | Source | Purpose |
|---|---|---|---|---|---|---|
| Aug 2, 2026 | Anton Makiievskyi | grantmaking.ai, Manifund | $50,000 | Interpretability | manifund | Getting Split-Personality Training ready for deployment |
| Aug 2026 | BlueDot Impact | $1,200 | Other | bluedot | Testing alignment faking as a defense against model poisoning | |
| Jul 17, 2026 | Foresight Institute | ~$150,000* | Epistemic infrastructure | foresight | Epistemic Memory for AI Reasoning Agents An AI reasoning agent is only as trustworthy as the memories it draws on. Florian Dietz is building an epistemic memory system for AI agents that tags every memory with its source and confidence, lets the agent trace where a claim came from and question it, and automatically swaps out unused memories while preserving the ones that matter. — AI for Science & Safety Nodes — 2026 cohort | |
| Apr 2026 | Foresight Institute | — | Other | grantmaking_ai | I received a seed grant from Foresight Institute. The seed has so far supported completing Phase 1 (prototype). This application seeks the follow-on funding needed to execute Phases 2–4 at realistic scale and speed. See details below. | |
| Jul 21, 2025 | Coefficient Giving | — | Other | grantmaking_ai | Coefficient Giving (formerly Open Philanthropy) funded the original research that resulted in the ICML paper with $175,000. | |
| May 16, 2025 | Austin Chen | Manifund | $1,000 | Alignment methods | manifund | Split Personality Training |
| May 16, 2025 | Marius Hobbhahn | Manifund | $2,000 | Alignment methods | manifund | Split Personality Training |
* Estimated typical grant size; Foresight Institute did not report grant-level amounts.