With GPT-5, OpenAI claims it’s their most advanced, fastest, and most powerful model to date – and it’s going to change the way businesses and individuals use AI. OpenAI describes GPT-5 as a big improvement over GPT-4o and the o-series, which were designed for complex thinking, advanced math, and systematic problem solving.
Big companies like BNY, Figma, Intercom, Lowe's, and Morgan Stanley have started to use the model. OpenAI says that its business products are now used by more than 5 million paying customers. This shows that a lot more businesses are using AI. At the same time, ChatGPT has grown to almost 700 million weekly active users, which shows how people have made this technology part of their everyday lives.
Business partners such as Amgen and Uber say they're seeing good early results. They praise the model's ability to understand complex situations and give better quality results. According to Amgen, for instance, internal testing revealed that GPT-5 is already improving scientific work by providing more dependable results than previous iterations.
Why businesses are paying attention
What’s new with GPT-5? The model has three new usage modes—Fast, Thinking, and Auto—so you can balance speed with deeper reasoning. For businesses, this means more flexibility in workflows, from quick content drafting to careful decision analysis.
Companies like Salesforce and Lowe’s see GPT-5 as a tool to boost productivity and spark new ideas. Experts think wider enterprise adoption will mean better teamwork, faster problem-solving, and smarter decision-making across industries.
OpenAI is launching GPT-5 Pro, a high-end version that will have better reasoning and more in-depth results. This will attract attention from research groups and critical industries where precision matters.
A bumpy beginning and customer pushback
The GPT-5 launch was not without its challenges, despite the hype. Many users expressed frustration, claiming the model seemed less intelligent than GPT-4 and didn't deliver the game-changing leap people had been waiting for since GPT-4 came out in 2023.
The AI's personality and speech pattern were partly to blame for the irritation. People said GPT-5's answers seemed too stiff and didn't have the friendly, chatty feel users had gotten used to. OpenAI responded by announcing changes to how it acted, adjusting ChatGPT to add warmer touches like “Good question” and “Great start,” while making it clear these weren't just empty compliments.
Technical issues also fanned the flames. OpenAI pushed back GPT-5's launch several times because of safety tests and computer power limits. When it hit the market, server overload and system errors made for a rough start for many users. OpenAI's boss, Sam Altman, owned up to the problems, saying, “We messed up big time on some parts of the launch.” He admitted that upgrading a product used by hundreds of millions in one day turned out tougher than they had thought.
Changes and future plans
To calm things down, OpenAI has brought back GPT-4 as a "Legacy models" choice for Plus subscribers to keep using. Users can now access other older versions, too. This shows a new approach where the company plans to give users more time to switch before retiring popular models.
Also, OpenAI has raised usage limits. Plus and Team subscribers can now send up to 3,000 messages per week in Thinking mode. The company has made GPT-5 available to enterprise and education customers, which expands its reach even more.
For now, businesses are moving ahead with adoption, and OpenAI is already course-correcting to meet user expectations. Whether GPT-5 ultimately earns its place as the defining AI of this era will depend on how quickly the company can balance ambition with usability.
