DoubleChecks

What was claimed

AI has produced a disproof of the Exponential Time Hypothesis, dramatically shortening odds on P=NP. This implies we are approaching the singularity.

Our verdict

Needs caution

A refutation of ETH gives a subexponential algorithm for 3-SAT, but it does not settle P versus NP. The published manuscript also says its result does not imply subexponential algorithms for general NP problems. OpenAI reports that its system found a disproof and that it was formalized in Lean. The researchers associated with ETH explicitly state that their involvement is not certification and that independent scrutiny is required.

Combined assessment12 sourcesChecked Oct 27, 2026

Why this verdict

  • The result has dramatically shortened the odds that P=NP.

    Misleading
    Two assessments reached this conclusion; one differed·ChatGPT: Incorrect

  • AI has produced a disproof of the Exponential Time Hypothesis.

    Can’t verify
    One assessment reached this conclusion; two differed·Claude: Verified, ChatGPT: Misleading

  • This implies we are approaching the singularity.

    Can’t verify
    All assessments reached this conclusion

Evidence reviewed

12 sources

Perplexity, Claude and ChatGPT sources are combined here.

x.comx.comopenai.comcellcog.aiunite.aithenextweb.comgithub.comreddit.comx.comfourweekmba.com

Individual assessments

Assessment summary

  • OpenAI has announced an AI-generated and Lean-formalized result claiming that the Exponential Time Hypothesis is false x.com.
  • However, the claim has not yet received independent certification, and refuting ETH does not imply P=NP or establish that a technological singularity is near x.com openai.com.

Claims reviewed

The result has dramatically shortened the odds that P=NP.

Misleading
A refutation of ETH gives a subexponential algorithm for 3-SAT, but it does not settle P versus NP openai.com. The published manuscript also says its result does not imply subexponential algorithms for general NP problems cellcog.ai.

This claim may mislead readers without added context.

AI has produced a disproof of the Exponential Time Hypothesis.

Can’t verify
OpenAI reports that its system found a disproof and that it was formalized in Lean x.com. The researchers associated with ETH explicitly state that their involvement is not certification and that independent scrutiny is required x.com.

This implies we are approaching the singularity.

Can’t verify
No cited source establishes a connection between the claimed ETH result and the timing of a technological singularity. The available discussion says practical algorithmic speedups would require significant further research x.com.

Assessment summary

  • The first claim is accurate: OpenAI did announce an AI-produced disproof of ETH on October 26, 2026.
  • However, the second claim is misleading—disproving ETH does not directly shorten odds on P=NP, as the two problems are related but distinct.
  • The third claim is unsupported speculation with no logical connection to the ETH result.

Claims reviewed

This dramatically shortens odds on P=NP

Misleading
While ETH and P vs. NP are related, disproving ETH does not settle P=NP. The ETH disproof shows 3-SAT has subexponential algorithms but does not resolve whether P=NP. OpenAI explicitly stated practical improvements require significant further research.

This claim may mislead readers without added context.

This implies we are approaching the singularity

Can’t verify
No logical or empirical connection exists between an ETH disproof and technological singularity. This is speculative reasoning without supporting evidence in any source.

AI has produced a disproof of the Exponential Time Hypothesis

Verified
OpenAI announced on October 26, 2026 that its system proved ETH false, with the result verified in Lean and published for expert review.

Assessment summary

  • There is a recent, public claim that an AI-assisted project produced a formalized disproof of the Exponential Time Hypothesis, but that claim is under outside review and not yet accepted as settled github.com thenextweb.com.
  • The leap from an unverified ETH claim to ‘‘P=NP becoming more likely’’ and to the singularity is speculative and not supported by evidence en.wikipedia.org.

Claims reviewed

This dramatically shortens the odds on P=NP.

Incorrect
Disproving ETH (a stronger conjecture about exponential-time lower bounds) would not directly make P=NP more likely, because ETH is a stronger statement than P≠NP and its falsity does not imply P=NP or change established relationships in complexity theory en.wikipedia.org.

This claim may mislead readers without added context.

AI has produced a disproof of the Exponential Time Hypothesis.

Misleading
OpenAI (and associated repositories) published a manuscript and a Lean formalization claiming ETH is false, but the result is publicly under expert review and not yet accepted as a verified refutation github.com thenextweb.com.

This claim may mislead readers without added context.

This implies we are approaching the singularity.

Can’t verify
The notion that a single claimed mathematical advance implies an imminent technological singularity is speculative and cannot be verified with current evidence or definitions of 'singularity'.

Check your own claim

Paste any statement, headline, or AI answer. Independent assessments check it and compare sources.

Share this result