How confessions can keep language models honest

OpenAI researchers are testing a method called 'confessions' that trains models to admit when they make mistakes or act undesirably, aiming to improve AI honesty, transparency, and trust.

From the source
OpenAI researchers are testing “confessions,” a method that trains models to admit when they make mistakes or act undesirably, helping improve AI honesty, transparency, and trust in model outputs.
openai.com

Visit source ↗openai.com

Announced: 03 Dec 2025 10:00 UTC
Detected: 03 Aug 2026 21:11 UTC (+243d after)

Other stories

OpenAI · 28 AugOur decision on Cursor following its acquisition by SpaceX

OpenAI · 28 AugSupporting Thailand’s next generation of AI startupsPartnership

Google Research · 27 AugPlanetary prediction engine: Automating global models via Earth AIResearch paper

OpenAI · 27 AugBetter answers, broader thinking: What students gain from ChatGPT and critical-thinking training

Research paper

OpenAI · 27 AugExpanding OpenAI’s presence in BrazilAvailability change

LG AI Research · 27 Aug[ACL 2026] ReSQL: A Self-Improving Text-to-SQL Framework through Retrieval-Augmented Error ReasoningResearch paper