AI news story

New math benchmark reveals AI models confidently solve problems that have no solution

A consortium of 64 mathematicians built SOOHAK, a new AI benchmark with 439 handwritten tasks, including 99 that are deliberately unsolvable. Google's Gemini 3 Pro leads on research-level problems at 30 percent. But no model cracks 50 percent on spot

  • LLMs
  • Source: The Decoder
  • Published: 2026-05-17
  • Signal score: 4
  • 96 sources

Editor's take

A new benchmark, SOOHAK, designed by mathematicians, demonstrates that current large language models confidently provide solutions to intentionally unsolvable mathematical problems.

This revelation is significant because it exposes a critical blind spot in AI's reasoning capabilities, particularly for advanced models like Google's Gemini 3 Pro, which still falters on these fabricated challenges. It highlights the gap between pattern recognition and genuine logical deduction, impacting the reliability of AI in fields requiring absolute mathematical certainty.

Future research should focus on how models are trained to identify or flag unsolvable problems, rather than simply generating plausible-sounding, but incorrect, answers. Understanding the failure modes of models like Gemini 3 Pro on SOOHAK will be crucial for developing more robust and trustworthy AI systems.

Signal score: 4

This event was corroborated by 96 independent sources. The signal score weighs cross-source corroboration, recency, source weight and topic salience. How we rank stories.

More LLMs stories

  1. OpenAI acquires presentation startup NextSlide

    TechCrunch · 2026-08-08

    NextSlide says its team members are now working on ChatGPT.

  2. Claude Vs ChatGPT: How These AI Assistants Differ

    Engadget · 2026-08-08

    In a practical breakdown of how Claude and ChatGPT AI models differ, one tends to fall short when it comes to quality responses and overall user experience.

  3. Anthropic sets Claude Code to Auto Mode by default to protect developers from bad approvals

    The Decoder · 2026-08-08

    Starting August 14, Anthropic will make Auto Mode in Claude Code the default for Pro, Max, and Team plans. The company says it's safer.

  4. Responding to the next frontier of critical cyber capabilities

    OpenAI Blog · 2026-08-07

    OpenAI is sharing preliminary cybersecurity evaluations for Astra and the steps we’re taking to strengthen safeguards and security controls.

  5. OpenAI says it slowed Astra model development over security concerns

    TechCrunch · 2026-08-07

    OpenAI said this model, which is still in development, reached its "critical cybersecurity threshold," meaning it could independently identify and carry out cyberattacks against

  6. Presentation: Keeping ChatGPT Fast as AI Development Accelerates

    InfoQ · 2026-08-08

    Martin Spier explains how agentic workflows dramatically increase code change volume at OpenAI. He d