OpenAI’s Sam Altman urges ‘code red’ ChatGPT push as rivals move in

NEW YORK – OpenAI chief executive officer Sam Altman is redirecting internal resources to speed up improvements to ChatGPT, declaring a “code red” situation that will delay work on other initiatives, according to a person familiar with the matter. 

Mr Altman on Dec 1 sent an internal memo calling for a “surge” to improve ChatGPT, said the person. Mr Altman also called for staff to delay progress on other efforts such as autonomous AI agents and advertising, the Information reported initially, citing the memo.

While Mr Altman did not say exactly what fixes needed to be prioritised and why, the Information said he recently warned employees that Google’s AI resurgence could cause temporary economic headwinds for OpenAI.

The report underscores how intense the rivalry among major AI developers has become as the entire industry works to convince Wall Street that the billions of dollars they have invested in the technology will turn a profit.

Alphabet’s Google released the AI model Gemini 3 in November that gained almost immediate praise for its capacity to reason and code, as well as to perform niche tasks that have tripped up AI chatbots.

OpenAI took Google head-on in October by unveiling its first AI-powered web browser.

Mr Nick Turley, OpenAI’s head of ChatGPT, reinforced the prioritisation of the chatbot in an X post late on Dec 1, saying the company’s focus is now “to keep making ChatGPT more capable, continue growing and expand access around the world – while making it feel even more intuitive and personal”.

As part of his push to prioritise ChatGPT improvements, Mr Altman has encouraged temporary team transfers and planned for a daily call with those responsible for the work, the Wall Street Journal reported.

Mr Altman said in an internal Slack memo that he wanted staff to improve features of ChatGPT so the user experience could be more customised, according to the Information. He added that he wanted to focus on improving AI model behaviour and image-generation capabilities, the news outlet reported.

The Information separately reported on Dec 2 that OpenAI has been working on a new AI model code-named “Garlic” that was performing well on tasks involving coding and reason. 

Google said in November that 650 million people were using its Gemini app. OpenAI recently said ChatGPT hit 800 million weekly users.

As of October, Gemini’s app had 73 million monthly downloads, well shy of ChatGPT’s 93 million monthly downloads, according to research firm Sensor Tower.

OpenAI rival Anthropic PBC rolled out a new version of its AI model in November that is designed to be better at automating coding and office tasks, intensifying its own competition with OpenAI and Google to win over business customers. BLOOMBERG