“MMonopoly is Silicon Valley’s answer to Darth Vader and is “a condition of all successful business,” said Peter Thiel. This aspiration is widely shared by Valley giant Gamman, his new acronym for Google, Apple, Microsoft, Meta, Amazon, and Nvidia. And with the advent of AI, each one’s desire to reach that blessed state before others gets there is even greater.

One sign of their anxiety is that they are spending insane amounts of money on the 70-odd generative AI startups that have proliferated since it became clear that AI was going to be the new thing. Microsoft, for example, reportedly spent $13bn (about £10.4bn) on OpenAI, while leading a $1.3bn funding round for DeepMind co-founder Mustafa Suleiman’s startup Inflection. He was also an investor. Amazon invested $4 billion in Anthropic, a startup founded by refugees from OpenAI. Google invested $500 million in the same business, he pledged an additional $1.5 billion, and he invested an unknown amount in A121 Labs and Hugging Face.
(Yes, I know the name doesn’t mean anything.) Microsoft also invested in his French AI startup, Mistral. and so on. In 2023, only $9 billion of the $27 billion invested in AI startups was invested. From a venture capitalist company –Until recently, the company was by far the largest funder of emerging technology companies in Silicon Valley.

what’s happening? After all, the big tech companies have their own “fundamental” AI models and don’t need what smaller companies have built or are building. And every penny drops. We’ve seen this strategy before. An existing company discovers and captures potential competitors at an early stage. For example, Google acquired YouTube in his 2006. Facebook acquired Instagram for $1 billion in 2012 when it had only 13 employees, and WhatsApp in 2014 (for $19 billion, which seemed an exorbitant amount at the time).

With the 20/20 vision of hindsight, we now see that these were all anti-competitive acquisitions that should have been resisted at the time and were not. That’s why it’s so refreshing to know that at least one regulator, the UK’s Competition and Markets Authority (CMA), seems determined to learn from its history.

in Speech given at a gathering of American antitrust lawyers Just over a week ago in Washington, CMA CEO Sara Cardel called for ensuring the market for fundamental AI models is supported by fair, open and effective competition and strong consumer protections. announced that he had decided to do so. Her concern is that the growing presence of a few large incumbents across the AI ​​value chain (the series of steps required to turn inputs into usable outputs) will undermine competition and limit companies’ options. This meant that there was a possibility that these markets could be formed in a way that degraded quality. and consumers.

She cited three major risks to competition. One is that companies that control critical inputs for developing the underlying model may restrict access to protect themselves from competition. Powerful incumbents may exploit their positions in consumer and business markets to limit competition in model deployment and thereby distort choice. And we believe that partnerships between key players have the potential to strengthen or expand existing market power across the value chain.

He also said the CMA would take action to assess and mitigate competition risks from new technologies through its formidable investigatory powers, including merger control reviews, market investigations and possible designations under new digital competition laws. I warned you.

It was truly amazing to hear a major regulator speak like this about the technology industry. Cardel said the CMA will be a technology industry that believes in being proactive and (as is often said) moving quickly to break things, rather than waiting for problems to arise before acting. He suggested that he would try to stay ahead of the big players rather than lag behind them. He said the CMA is already preparing for this task based on what it has learned so far from adapting to technology platforms. Rather than focusing only on individual parts of the chain, the value of AI model deploymenthe aims to look at the entire chain holistically. It also plans to use its merger review powers more aggressively to assess the impact of alliances and AI investments on competition.

Isn’t that exciting? But in some ways it is no surprise as it is one of the few British institutions that seems able to use the post-Brexit freedoms as an opportunity for creativity and innovation. And bigwigs who are tempted to dismiss Cardel’s speech as mere fiery rhetoric should reflect on the CMA’s recent track record. A thorough investigation into Microsoft’s acquisition of Activision Blizzard, for example; or how Meta forced the sale of Giphy, an online database and search engine that allows users to find and share animated GIF files. Cardel may be lower profile than her U.S. FTC counterpart Lina Khan, but it’s clear she means business. People with strong possessiveness should be careful.