#Safety Articles

10 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 08/25/2026 Read AI Rating: Medium

OpenAI Bans Russia-Linked ChatGPT Accounts Behind Fake 'Think Tank' Influence Campaign

OpenAI disrupted a covert influence operation, very likely run from Russia, that used ChatGPT to build a fake Israel-based think tank and a pro-Russia 'sovereignty index.'
ChatGPT News 08/18/2026 Read AI Rating: High

Introducing ChatGPT for Teens: Built for Learning, Backed by Protections

OpenAI is launching ChatGPT for Teens, a dedicated experience for ages 13-17 with default safety guardrails around self-harm, sexual content, and romantic roleplay, plus a Study Mode-centered learning design and parental controls.
Claude News 08/14/2026 Read AI Rating: Medium

How Claude's Text Watermarking Works

Anthropic will start watermarking text from future Claude models using SynthID-Text, a technique that biases word-choice randomness in a detectable but statistically invisible way, driven primarily by the EU AI Act's transparency requirements for AI-generated content.
Claude News 08/07/2026 Read AI Rating: Medium

Anthropic Cuts Claude Fable 5's Biology False-Positive Refusals by 85%

Anthropic rewrote the safety classifier 'constitution' governing when Claude Fable 5 routes biology-related queries to the less capable Opus 5, cutting unnecessary fallbacks by roughly 85% while still blocking genuinely dangerous dual-use requests.
ChatGPT News 07/31/2026 Read AI Rating: Low

OpenAI Details Its Responsible AI Practices Across Europe: Safety, Security, Transparency, and Provenance

OpenAI has published an overview of how its safety, security, transparency, and content-provenance practices align with the EU's AI governance framework, including the EU AI Act's codes of practice.
ChatGPT News 07/31/2026 Read AI Rating: Medium

OpenAI Keeps Publishing "Disrupting Malicious Uses of AI" Reports โ€” 44-Plus Cases From State Influence Ops to Scam Networks

OpenAI has been publishing periodic "Disrupting malicious uses of AI" reports since 2024, repeatedly detecting and banning ChatGPT accounts tied to state-linked influence operations, scam networks, and malware-development assistance.
ChatGPT News 07/20/2026 Read AI Rating: Medium

Safety and Alignment in an Era of Long-Horizon Models

OpenAI shares safety and alignment lessons from a limited internal deployment of long-running models, describing misbehavior that pre-deployment evaluations failed to catch and the safeguards it introduced in response.
ChatGPT News 07/16/2026 Read AI Rating: Low

OpenAI Explains Age-Based Protections and Parental Controls for Safe Teen Access to AI

OpenAI has published an article outlining the age-appropriate protections, learning support features, and parental controls designed to help teenagers use ChatGPT safely.
ChatGPT News 07/15/2026 Read AI Rating: Low

OpenAI Proposes 'Reverse Federalism' Vision for Unified State-and-Federal AI Governance

OpenAI's Chris Lehane, Chief Global Affairs Officer, has published an article proposing an AI governance approach he calls 'reverse federalism,' combining state and federal law. He argues that safety legislation in leading states like California should serve as the foundation for federal-level alignment.