OpenAI released GPT-5, making it the new default model in ChatGPT for signed-in free and paid users. The company described it as a unified system, with a real-time router that decides when to answer quickly and when to reason at length.
What happened
- OpenAI said GPT-5 improved on earlier models in coding, writing, and health questions, and that with reasoning enabled its answers were around 80% less likely to contain a factual error than its o3 model.
- GPT-5 replaced GPT-4o and other older models in ChatGPT, prompting complaints from users, and within days OpenAI restored GPT-4o as an option for paid subscribers.
- OpenAI said ChatGPT was on track to reach around 700 million weekly users at the time of the launch.
- Competitors including Google and Anthropic released new frontier models in the following months.
Why it mattered
GPT-5 marked another step in the rapid spread of generative AI into everyday work and intensified competition among AI developers. It also raised fresh questions for organisations about staff using consumer AI tools with business data.
Lessons for organisations
When AI tools change or models are replaced, review outputs and controls again, as behaviour can shift. An AI management system such as ISO/IEC 42001 helps keep track of where AI is used and who checks its results.
Part of our Top stories archive of headline-making events in information security, privacy, and AI. If you would like help applying the lessons to your organisation, contact us.
