Microsoft is trying to reduce its reliance on OpenAI by developing a cheaper, less powerful AI model, report says

• Microsoft wants to break its reliance on OpenAI, per The Information.
• The decision is largely motivated by the cost of running advanced AI models, per the report.

Microsoft's partnership with OpenAI has given the company a new lease of life.

The surprise popularity of the company's chatbot, ChatGPT, lit a fire under one of the tech giant's oldest rivals — Google — and sparked an ongoing AI arms race in the tech world.

However, The Information reported that Microsoft was trying to reduce its reliance on the AI lab. The decision is largely motivated by the spiraling costs of running advanced AI models, the publication said.

One current employee and another ex-worker who recently left the company told The Information that an executive who oversees 1,500 researchers at Microsoft instructed some of them to develop an in-house conversational AI.

The company wants its in-house large language models to be cheaper and smaller than OpenAI's, even if this means they don't perform as well, the people told The Information.

Product teams were already working on incorporating the company's in-house AI programs into products such as Bing Chat, the people added.

Representatives for OpenAI and Microsoft did not immediately respond to Insider's request for comment, made outside normal working hours.

Microsoft hasn't been shy about its partnership with OpenAI. In 2019, the company invested $1 billion into the ChatGPT maker and reportedly put a further $10 billion toward the company after the release of the AI-powered chatbot.

The tech giant has incorporated the AI lab's tech into flagship products, including its search engine, Bing.