The UK Competition and Markets Authority (CMA) has cleared Amazon’s artificial intelligence (AI) partnership with Anthropic, concluding that the agreement does not constitute a relevant merger situation under UK competition law. Announced last year, the partnership saw the e-commerce giant invest some $4bn into Anthropic and include a non-exclusive arrangement for Amazon Web Services (AWS) to supply compute resources to the AI startup.

As part of the collaboration, Anthropic is leveraging AWS as its primary cloud provider for specific workloads. Additionally, the two parties have committed to partner on the development of future Trainium and Inferentia technology.

Moreover, Anthropic has made a non-exclusive, long-term commitment to provide access to its models on Amazon Bedrock, while Amazon has secured a non-exclusive licence to use Anthropic’s FMs in its services.

Amazon has also obtained certain rights related to its investment, including consultation rights on significant business issues at Anthropic. Additionally, it has a right of first notification regarding any potential change of control over Anthropic, allowing Amazon to submit a counteroffer within a specified time.

Anthropic announcement latest example of blunted CMA AI probe

Established in 2021, Anthropic specialises in AI safety and research and has developed the Claude family of FMs. The AI firm has raised capital from various investors, including Alphabet, Menlo Ventures, and Spark Capital.

The phase I inquiry into the AI partnership between Anthropic and Amazon was launched by the CMA in August 2024. The British competition watchdog reviewed whether the partnership might give Amazon material influence over the AI firm, which could meet the criterion of the enterprises ceasing to be distinct.

However, the CMA found that Anthropic’s UK turnover does not exceed £70m, nor do the parties, based on available evidence, jointly account for a 25% or more share of the supply of any goods or services in the UK.

As a result, the regulator concluded that the partnership did not qualify for further investigation under merger control regulations and opted not to refer the case under section 22 of the Enterprise Act 2002.

The antitrust regulator’s latest decision comes shortly after its clearance of another partnership in the AI sector between Microsoft and the machine learning and generative AI company Inflection AI.

In that case, the CMA similarly found no evidence of competition concerns after a phase 1 investigation, which concluded that the partnership did not create harmful effects in the UK market. A similar probe into Mistral was also discontinued. However, the UK CMA’s scrutiny of Alphabet’s partnership with Anthropic is still ongoing.

Read more: CMA delays UK cloud market report deadline to August 2025