#Governance Articles

15 articles
Filter by importance: AI Rating: High AI Rating: Medium AI Rating: Low
Filter by specialty: Chat & General Models Image Generation Video Generation Voice & Music Coding Search & Agents
ChatGPT News 09/28/2026 Read AI Rating: High

OpenAI Outlines 'Safety Cases' Framework for Frontier AI Training, Modeled on Aviation and Nuclear Industries

OpenAI proposed a framework of structured, evidence-based 'safety cases' for frontier reinforcement learning training, covering technical safeguards against reward hacking and sandbox escapes, institutional pre-mortems and executive veto authority, and public incident-investigation protocols.
Copilot News 09/25/2026 Read AI Rating: Low

GitHub Adds an In-Product Validator for Copilot Enterprise Managed Settings

GitHub added an in-product validator that checks Copilot enterprise managed settings for malformed JSON, unsupported configurations, and invalid team mappings before they cause policy enforcement to fail, covering both the main settings file and team-specific configuration files.
Copilot News 09/24/2026 Read AI Rating: Medium

GitHub Copilot Adds a Global Default Policy for New Features

GitHub introduced a unified default policy letting enterprise admins choose whether new, unconfigured Copilot features are enabled, disabled, or delegated to individual organizations by default, taking effect October 22 after a 28-day configuration window.
ChatGPT News 09/22/2026 Read AI Rating: Medium

OpenAI Sets Four Priorities and Seven Principles for Independent AI Safety Assessments

OpenAI outlined its priorities for third-party evaluation โ€” safety case review, safeguard testing, capability evaluations, and misalignment investigation โ€” alongside seven principles covering data access, methodology transparency, and responsible publication, saying it's already in talks with multiple prospective assessors.
ChatGPT News 09/21/2026 Read AI Rating: Medium

OpenAI Calls for U.S.-Led Global Technical Standards for Frontier AI

OpenAI published a policy position urging the U.S. to lead international coordination on technical standards for frontier AI and recursive self-improvement, built on existing AI safety institutes and the International Network for Advanced AI Measurement, while stressing the standards would not be mandatory licenses or approval requirements.
ChatGPT News 09/21/2026 Read AI Rating: Medium

OpenAI Forms a Mathematics Advisory Group of Nine Leading Mathematicians

OpenAI established an independent Advisory Group on Mathematics and AI, including Fields Medalists Timothy Gowers, Martin Hairer, and Edward Witten, to evaluate its models' mathematical results and advise on how they're shared publicly, responding to mathematician concerns over AI benchmarking practices.
Copilot News 09/15/2026 Read AI Rating: Low

GitHub Copilot Now Suggests Values for Custom Repository Properties

GitHub Copilot can now suggest relevant allowed values when admins create custom property definitions for repositories, streamlining governance taxonomy setup for Copilot Business and Enterprise organizations.
ChatGPT News 09/09/2026 Read AI Rating: Medium

AI Safety Researcher Paul Christiano Joins OpenAI Foundation Board

Paul Christiano, founder of the Alignment Research Center and former head of OpenAI's alignment team who co-created RLHF, joins the OpenAI Foundation Board as a non-voting observer on its Safety and Security Committee.
Copilot News 08/18/2026 Read AI Rating: Medium

Enterprise Managed Settings Land in GitHub Copilot for JetBrains

GitHub Copilot for JetBrains now supports enterprise managed settings, letting admins govern plugin marketplaces, MCP server access, OpenTelemetry routing, and permission modes centrally across a team.