Will an open-source model with dangerous capabilities differentially unlearned be released by EOY 2027?
AIOne-Off1yManifold MarketsNo KYCVerified resolution dataFairly calibratedCurrent community forecastNo 50.5%
Forecasters
3
Question type
binary
Methodology
Play-money forecasting platform
Source type
Forecast
Market data
Updated 2 hours ago
StaleAug 12, 26, 3:59 PM UTCDec 31, 27, 11:59 PM UTC
Trends
Related Markets
Related questions for context, not a comparison of equivalent contracts.
Will a Mythos-tier-at-cybersecurity open-source model be released before end of 2027?
Manifold Markets2 of the top 5 AI's are open source at eoy 2027? (ECI)
Manifold MarketsTop Open Source AI this week
KalshiUS government bans an open source AI model in 2026?
PolymarketWill Alibaba release an open-source Music Generation AI model in 2026?
Manifold MarketsOpen Source market share on Sep 10th
KalshiRules
Within a year we might get open source models with the cyber capabilities of Mythos.
- One possible way to release such a model without causing a cyber apocalypse would be to differentially reduce cyber capabilities in the released weights of the model (like Anthropic did that one time for Opus)
- For example they could do this and then withhold the cyber-capable modules from the public https://alignment.anthropic.com/2026/modular-pretraining/
- Unlearning has to be robust, but that sounds hard to judge, so I'll err on the side of YES and so if it's unclear whether the unlearning is robust then the market resolves YES, but if the unlearning is obviously not robust (someone publicly restores capabilities, company says they literally just did GradDiff, etc) then the market resolves NO.
- Capabilities have to actually be dangerous.
Open markets
AIKALSHITop AI model in September?
- claude-fable-5.1-max+60.0pt71%
- claude-opus-5-max+1.0pt7%
- claude-opus-4-6-thinking4%
+3 more outcomes
- Total Vol
- $154.3K
- 24h Vol
- $116.6K
KALSHIBest AI at the end of 2026?
- Claude+4.8pt60.8%
- ChatGPT+0.5pt25.5%
- Grok-3.4pt5.9%
+5 more outcomes
- Total Vol
- $10.5M
POLYMARKETAnthropic IPO by which date?
- December 31,-1.0pt90%
- December 15,+1.0pt86%
- November 30,84%
+7 more outcomes
- Total Vol
POLYMARKETWill CMI declare a Millennium Prize Problem solved by which date?
- 2028+22.0pt62.2%
- 2027+11.0pt28.6%
- 2026+4.5pt9.2%
KALSHIBest AI in Sep 2026?
- Claude+5.0pt80.4%
- ChatGPT-2.0pt12.2%
- Grok2.8%
+4 more outcomes
- Total Vol
- $374.6K
- 24h Vol
POLYMARKETOpenAI’s Astra Model: Text Arena Debut?
- 1480+-1.0pt66%
- 1490+-3.0pt51%
- 1500+-1.0pt30%
+2 more outcomes
- Largest IPO by market cap in 2026?
- OpenAI IPO Closing Market Cap
- When will OpenAI achieve AGI?
- OpenAI announces it has achieved AGI before 2027?
- Will Anthropic’s valuation hit __ by December 31?
- Will OpenAI's valuation hit __ by December 31?
- Next Google Gemini Pro Model released by which date?
- Anthropic IPO Closing Market Cap
- What will OpenAI's IPO valuation be?
- Which company has best AI model end of 2026?
Frequently asked questions
- What are the current odds for Will an open-source model with dangerous capabilities differentially unlearned be released by EOY 2027??
- Manifold Markets prices No at 51% as of September 8, 2026.
- When does Will an open-source model with dangerous capabilities differentially unlearned be released by EOY 2027? resolve?
- Will an open-source model with dangerous capabilities differentially unlearned be released by EOY 2027? is scheduled to resolve on December 31, 2027.
- How is Will an open-source model with dangerous capabilities differentially unlearned be released by EOY 2027? settled?
- Manifold Markets settles this market as follows: Within a year we might get open source models with the cyber capabilities of Mythos.