Altman says AI benefits outweigh some harms, while rejecting catastrophic risks.
OpenAI is merging Codex into the ChatGPT desktop app, adding a Work setting and releasing GPT-5.6 models as it competes to become a core tool for professionals.
OpenAI announced that Codex, its AI coding tool, is being merged into the ChatGPT desktop app. The move positions ChatGPT as a broader work hub for engineers and other professionals rather than a standalone chatbot experience. OpenAI also introduced a Work setting for ChatGPT alongside the release of GPT-5.6 models.
The release reflects OpenAI’s strategy of training powerful AI models and delivering them through ChatGPT to a large user base. OpenAI engineer Thibault Sottiaux said on a Thursday livestream that ChatGPT has almost 1 billion users. For workplace users, the key takeaway is that OpenAI wants ChatGPT to handle more of the tools and workflows professionals already use.
Codex began as a software engineering tool, but OpenAI says its use has broadened. A June company report said data analysis and research were growing faster than traditional coding uses on Codex, and that 5 million people use the tool each week. That shift helps explain why OpenAI is bringing Codex capabilities into ChatGPT instead of keeping them separated.
OpenAI demonstrated a new Work feature for the ChatGPT desktop app that incorporates popular Codex capabilities. Those include the ability to modify files on a computer and operate autonomously in a browser. Codex leader Andrew Ambrosino described the ChatGPT merger as “only the first” step toward unifying the experience across web, mobile, and desktop.
OpenAI is competing release by release with Anthropic while also trying to stay ahead of Google, Meta, SpaceXAI, and cheaper Chinese AI models. Meta also announced its first paid AI model on Thursday. Business Insider reported that OpenAI has filed confidential paperwork to go public, but when CNBC asked Sam Altman if OpenAI would go public this year, he said, “I don’t know.”
Altman says AI benefits outweigh some harms, while rejecting catastrophic risks.
A former OpenAI safety leader says the company is not being careful enough.

A departing OpenAI safety employee says the company’s culture is broken and calls for stronger safeguards.

A former OpenAI safety researcher says AI companies need deeper safeguards, not trial-and-error risk management.