Amazon and artificial intelligence research firm Anthropic are being investigated as a de facto merger in the United Kingdom.
The UK’s Competition and Markets Authority (CMA) issued a formal notice on Aug. 8 announcing the inquiry’s commencement. According to documentation published on the UK government website, this marks the beginning of a “phase 1” investigation.
The merger question
As Cointelegraph previously reported, UK authorities announced they were conducting preliminary investigations into partnerships between Amazon and Anthropic, Microsoft and Mistral AI, and Microsoft and OpenAI to determine whether any of the relationships had run afoul of EU regulations.
CMA documents published in April indicate UK regulators are looking into Amazon’s $4 billion investments in Anthropic as well as their exclusivity agreements.
The current phase 1 inquiry officially begins on Aug. 9. The CMA says it will announce whether it intends to conduct a more in-depth “phase 2” of the investigation on Oct. 4.
At the heart of the investigation lies the question of whether Amazon’s minority ownership and the two company’s exclusivity amounts to an anticompetitive partnership.
It’s possible that regulators could view Amazon’s $4 billion investment and minority ownership as the motivating factor in Anthropic’s decision to exclusively use Amazon cloud services. While this doesn’t necessarily define a “merger” situation, antitrust regulators could simply be performing due diligence to ensure UK antitrust laws are being obeyed.
Big tech and AI
Companies in the big tech elite category are largely known for gobbling up smaller firms, especially in the field of AI. Google’s purchase of DeepMind was one of the most impactful in recent years as was its acquisition of Fitbit. And both Microsoft and Amazon have bought hundreds of companies over the past few decades.
But there are some key differences when it comes to the Amazon/Anthropic and Microsoft/OpenAI partnerships when compared to the typical big tech buyout.
When Google parent company Alphabet bought DeepMind, for example, most regulators’ concerns appeared to center around the fact that the AI firm had partnered with the UK government on numerous projects involving private citizens’ healthcare data. While the deal faced some anti-trust scrutiny, ultimately it was determined that Alphabet was simply purchasing one of its many competitors.
OpenAI and Anthropic, however, have a different business model than DeepMind, FitBit, or just about any other big tech partnership featuring investments in excess of $1 billion resulting in an ownership stake.
Both AI firms claim their business model is dedicated to developing a human-level AI system for the purpose of ensuring machines don’t rise up and destroy us. As industry leaders in the field of generative AI, Anthropic and OpenAI are, arguably, foundational companies in the field.
Ostensibly, if a company were to purchase OpenAI or Anthropic outright in a complete takeover and buyout, it could be seen by regulators as a move to keep the generative AI technology out of competitors' hand as a means to monopolize the chatbot market.
While neither Amazon nor Microsoft have purchased their respective partners in Anthropic and OpenAI, the question for regulators is whether their “strategic partnerships” amount to the same form of anti-competition under a different name.
In Microsoft’s case, it’s gone so far as to file official paperwork indicating that it considers OpenAI a competitor, despite having invested at least $13 billion in the AI firm and having previously held a seat on its board of directors.
Related: Speculation runs wild for new GPT model after Altman posts strawberry garden