AIAI Programs Hub
All programs / Anthropic Model Safety Bug Bounty

Bounty · Anthropic

Anthropic Model Safety Bug Bounty

Model safety bounty for verified universal jailbreaks in high-risk CBRN/cyber domains

watch / prior and expanded bounty rounds yes, online remote

Historical record: details were last verified on 2026-07-08. Confirm current conditions with the official organizer.

Program overview

Strong AI safety red-team target if Anthropic opens another round

Notes

Do not confuse with Anthropic HackerOne infrastructure program or disclosure-only cyber-jailbreak page

View official program ↗

Quantiacs Contest 25 · Quant research challenge

Compete for $2M in strategy allocations.

Submit a unique trading system. The seven highest Contest Scores receive one-year notional allocations with performance fees currently based on a 10% profit share.

1st$1M
2nd$500K
3rd$250K
4th$100K
5th–7th$50K each

Unique systems only. Quantiacs applies a correlation filter, and allocations may be reduced or eliminated if a winner’s live Contest Score falls below its score at contest close.

Enter Quantiacs Contest 25 ↗ Official Q25 rules

Sponsored registration link. Prize amounts are notional strategy allocations; payouts are performance fees under Quantiacs’ Relative Performance Model. Winner allocations commence within 30 days after the contest ends. Eligibility and official rules apply; no purchase is necessary.