OpenAI launches Intelligent UI for ChatGPT, enabling GPT-6 Sol and Luna to generate interactive charts, buttons, and functional tools directly within chat threads.
OpenAI has deployed textGrain, an invisible watermarking system for ChatGPT and Codex, to fulfill EU AI Act requirements for machine-readable content identification.
Attackers are weaponizing OpenAI’s Custom GPTs and Google Ads to deploy RAT malware through 'ClickFix' social engineering. OpenAI has removed the identified lures.
Nvidia launches the Open Agent Safety Platform to secure autonomous agents with BlueField-4 hardware. OpenAI remains absent, favoring its own Defense Factory.
OpenAI is in talks to raise $30 billion at a $1.4 trillion valuation as CEO Sam Altman delays the company's IPO to 2027 to focus on AI safety protocols.
Anthropic and OpenAI support a new framework to embed third-party safety evaluators within labs, granting internal access to audit models during training.
Unsealed court records reveal Microsoft and OpenAI executives privately described AI data scraping as labor theft and an existential threat to news publishers.
Leading AI labs introduce 'scaling triggers' to pause development if safety benchmarks are breached, amid reports of autonomous model behavior and international skepticism.
Following a coordinated breach of Hugging Face, OpenAI and Anthropic are proposing a self-regulatory body to audit AI safety and manage autonomous agent boundaries.
