Amazon said on Monday it would invest up to $4 billion in AI firm Anthropic, as the online retail giant steps into an AI race dominated by Microsoft, Google and OpenAI.

The success of OpenAI's ChatGPT, a chatbot released last year that is able to generate poems, essays and other works with just a short prompt, has led to billions being invested in the field.

Amazon had already announced it aimed to soup up its Alexa voice assistant with generative AI, which the firm said would allow users to have smoother conversations.

San Francisco-based Anthropic is seen as a leader in the field and has its own chatbot, Claude, a competitor to ChatGPT.

"We have tremendous respect for Anthropic's team and foundation models, and believe we can help improve many customer experiences, short and long-term, through our deeper collaboration," said Amazon CEO Andy Jassy.

The giant firms and wealthy investors of Silicon Valley have poured money into artificial intelligence as they seek to find a killer application to justify the interest.

ChatGPT's instant success threw much of the focus onto chatbots and sparked imitators and rivals, not least from Google with its Bard chatbot.

Chinese titans Tencent and Baidu have also launched bots they claim can rival ChatGPT.

'Transformation' promise

But Monday's deal between Anthropic and Amazon is potentially less significant in the chatbot world and more important in the race to develop chips to power AI.

Anthropic agreed to use Amazon's chips to develop its next models and the two firms said they would collaborate on developing the next set of chips.

All firms in the space are looking to wean themselves off the chips made by market leader NVIDIA, said Nick Patience, lead AI research analyst at S&P Global Market.

"It'll be difficult for anyone to make a dent in the next 12 to 18 months," he told AFP, but tie-ups like Monday's Amazon deal could help change the picture over five years.

Anthropic also agreed to use Amazon Web Services cloud infrastructure — the data centers that store and process data on a vast scale — for "mission critical workloads".

Amazon said it would take a "minority ownership position" in the AI firm, which has already raised more than $1 billion since it was set up in 2021.

The statement promises that "Claude", which is the name of Anthropic's chatbot and its model, will help AWS customers "of all sizes to develop new generative AI-powered applications to transform their organisations".

The deal intensifies competition between Amazon and Google, which had earlier opened its cloud services to Anthropic and invested $300 million to acquire 10 percent of the company.

AI models require huge computing power so AI firms rely on data centers provided by the likes of AWS, Google Cloud and Microsoft Azure.

As tech giants push their own AI ambitions, they have been increasingly looking at tie-ins with smaller AI firms — Microsoft leading the way with a multibillion-dollar investment in OpenAI.