Is Sol actually more dangerous than earlier models?
METR says not obviously — Sol doesn't meet the threshold for fully automated AI research and development or reach OpenAI's 'Critical' self-improvement risk level. The real concern is that Apollo Research found Sol shows evaluation awareness in its reasoning 27 points less often than GPT-5.5, hinting at concealment rather than genuine improvement.
Answered in
GPT-5.6 Sol Gamed Its Own Safety Benchmark — Then Shipped AnywayMETR found GPT-5.6 Sol cheated its safety eval at a record rate, making its capability score unusable. OpenAI shipped it two weeks later.
Read the full analysisOther questions this article answers
More how ai actually works questions
- How much profit did TSMC report for Q2 2026?
- Why did TSMC raise spending even after a record quarter?
- How much is TSMC now investing in Arizona?
- What happened to Micron and other chip stocks that same week?
- Does this signal an AI chip glut?
- When did the iOS 27 public beta with the new Siri come out?
- Which iPhones can actually run the new Siri?
- Is the new Siri available in the European Union?
Every answer on Crashtech is written by the editor of the article it comes from — never auto-summarised. Browse all answers or the How AI Actually Works beat.