OpenAI rewrites its AI model guidelines around five values
The Model Spec is OpenAI's public document setting out how its AI models should behave, and this revision reorganizes it.
By Standard Topic. AI-generated from the reporting listed under Coverage. Published . Updated .
This story is no longer being updated. It reflects reporting up to .
What happened
OpenAI released a revised Model Spec, a public roadmap of how its AI models should behave, reorganized around five core values. The values include faithfulness to human intent, harm minimization, honesty and transparency, human agency, and care. The company also acknowledged that its current models fall short of the framework.
Why it matters
The revision sets out how GPT models should use conversation history to work out what users want and disclose their assumptions. Commentators note these models have already proven hard to control.
Timeline
- Oct 29, 2:34 PM ET: OpenAI updated its Model Spec, reorganizing its AI behavior framework around five core values.
- Oct 30, 9:43 AM ET: OpenAI said its current models fall short of the updated framework, with faithfulness to human intent placed first among the values.
Coverage
- Five new principles are shaping the future of ChatGPT... OpenAI is rebuilding the rules of AI behavior (سيد بدري)
- OpenAI sets new behavioral guidelines for AI models that have already proven hard to control (the-decoder.com)
- OpenAI Rebuilds Model Spec Around Five Core Values (technology.org)
More from Technology & AI
- China puts humanoid robots to work serving food and sorting parcels
- Newsom proposes sweeping AI safety rules for Washington
- Google releases Gemini 4 Neon AI model with million-token context
- Microsoft and Intel widen their cloud chip partnership
- Google DeepMind rolls out Genie 4 to select subscribers
- AI helped AMD shorten chip design work, OpenAI says