Amazon in £3bn punt on artificial intelligence start-up

Amazon has become the latest tech giant to join the AI ​​arms race by pouring billions into the start-up behind a popular chatbot.

The online shopping company said it will invest up to £3.3 billion in San Francisco-based Anthropic, the developer of AI-powered chatbot Claude, a competitor to OpenAI’s ChatGPT.

Amazon will initially invest £1 billion for a minority stake in the artificial intelligence (AI) security and research company, with an option to increase the cash injection to the full amount later.

Amazon CEO Andy Jassy said, “We have tremendous respect for Anthropic’s team and grassroots models and believe that through our deeper collaboration we can help improve many customer experiences, short and long term.”

The price tag is likely to significantly increase Anthropic’s valuation, following a fundraising earlier this year that valued the entire company at almost £4.1 billion.

Claire Holubowskyj, senior research analyst at media research service Enders Analysis, said: “Amazon working with other AI companies isn’t new, but this is a big step forward.

“Amazon approaches AI at all levels, from the chips to the software, so this new investment simply means they’re putting their money where their mouth is.”

Under the terms of the deal, Anthropic will use Amazon Web Services (AWS) cloud computing system and AI computer chips from Amazon to create its programs.

It’s also a marked shift for Anthropic away from rival tech giant Google, which pumped £246 million into the company last year.

Amazon’s investment comes amid a fierce battle between some of the world’s largest tech companies as they scramble to lead the booming market for AI.

Microsoft has already forged an alliance with OpenAI, while rivals such as Inflection AI and Canada’s Cohere have raised hundreds of millions of pounds in funding from other major tech groups.

Amazon hopes to capitalize on growing interest in generative AI, technology capable of generating text and images in a manner similar to humans.

Amazon is also pitching its Trainium and Inferentia computer chips as viable alternatives to the chips developed by industry leader Nvidia, which has seen its stock price soar this year as the AI ​​boom created huge demand for its products.

The move follows the creation of AWS’s Bedrock service, which allows customers to build AI applications on the company’s cloud infrastructure and already has a range of products including Claude. But Amazon’s investment is not an exclusive arrangement, unlike Microsoft’s deal with OpenAI, which saw the latter adopt the tech giant’s Azure platform as its sole cloud provider.

The terms are also thought to give Amazon a much smaller share than the 49% of OpenAI that Microsoft controls.

Anthropic hopes it can surpass its rival in Silicon Valley, where its founders worked before splitting from the company in 2021 amid a dispute over the company’s direction. Microsoft fired the starting gun into Big Tech’s AI battle when it signed its partnership with OpenAI earlier this year and invested billions to advance the technology.

A key area of ​​interest for the tech giant is using AI to improve its Bing search engine as it looks for ways to loosen Google’s iron grip on the market.

Facebook owner Meta is also jumping on the AI ​​bandwagon, with the company releasing its own chatbot models, Llama and Llama 2, in an attempt to compete with OpenAI and Anthropic.

In addition to the tech giants, AI has also attracted the attention of the pharmaceutical sector as a way to accelerate the development of new drugs.

Drug giants like GSK have already started using AI to make medical treatments more personalized for individual patients, which in turn can increase the effectiveness of drugs. But the rapid rise of AI – and its ability to create realistic images, videos and human-like text – has prompted regulators and campaigners against the spread of misinformation to sound the alarm.

Last week, the Competition and Markets Authority (CMA) warned that AI could make consumers more vulnerable to email scams and fake reviews when shopping online.

The regulator has published seven principles to regulate AI models, which aim to build more accountability and transparency for companies looking to use the technology.