The "Into the Unknown" benchmark tasks nine AI agents with tackling open research questions from U.S. national laboratories. None of the 135 submissions meets all requirements. A total of 114 papers fail the integrity check, with agents claiming to have conducted experiments that never took place. GPT-6 Astra performs best but still doesn't meet the experts' standards. The article Nine AI agents tried autonomous research on open questions from US national labs and none passed the experts' bar appeared first on The Decoder.
Research1 min read
Nine AI agents tried autonomous research on open questions from US national labs and none passed the experts' bar
The "Into the Unknown" benchmark tasks nine AI agents with tackling open research questions from U.S. national laboratories. None of the 135 submissions meets all requirements. A total of 114 papers fail the integrity check, with agents claiming to have conducted experiments…
The DecoderBy: Jonathan KemperNovember 7, 2026 (34 minutes ago)
Available in 4 languagesAI Summary & Key Takeaways
- The "Into the Unknown" benchmark tasks nine AI agents with tackling open research questions from U.S
- national laboratories
- None of the 135 submissions meets all requirements
Research
KZ·AI 9513Инсайт дня
Nine AI agents tried autonomous research on open questions from US national labs and none passed the experts' bar
From the original source: The Decoder
View originalOpen original
Related stories
OpenAI lands an $11 billion funding round, but has to court dentists for more capital
OpenAI has secured $11 billion in fresh capital, the first tranche of a round that values the company at $1.25 trillion. The full round could swell to $20 billion. That OpenAI is already turning to wealthy private investors hints at a saturating AI capital market and growing…
The Decoder
OpenAI adds support for outbound phone calls to GPT-Live API. 📞
OpenAI adds support for outbound phone calls to GPT-Live API. 📞 Now developers can create voice AI agents that independently dial phone numbers and converse with people in real time. For example, an agent could call a restaurant and…
Telegram — Вайб-кодинг
OpenAI raises $11 billion at a $1.25 trillion valuation
✔️ OpenAI raises $11 billion at a $1.25 trillion valuation. The company successfully closed the first tranche of a new funding round, receiving $11 billion in its accounts. The total investment is expected to grow to $20 billion in the coming weeks.…
Telegram — AI Machine Learning Big Data
American sues OpenAI, accusing ChatGPT of helping plan his poisoning
⚖️ An American has filed a lawsuit against OpenAI, accusing ChatGPT of assisting in planning his poisoning. North Carolina resident James Gates claims his wife used ChatGPT to search for information on toxic substances and ways to conceal the poisoning. According to the man, he…
Telegram — GPT Time | Нейросети | ИИ