Artificial intelligence, US antitrust investigations on Nvidia and Microsoft

The US antitrust authorities are laying siege to artificial intelligence and its corporate leaders. The Federal Trade Commission has triggered a new federal investigation into the deal between Microsoft and one of its most valuable satellites in the technological universe of the future, Inflection AI. The intervention is in response to suspicions that Satya Nadella's group has artfully structured the deal with the aim of evading regulatory scrutiny.

The FTC and the Department of Justice, which share anti-monopoly responsibilities, have at the same time set down on paper the division of the tasks of supervision and investigation into AI in order to make their action more effective: the ministry will lead the examination of the business practices of Nvidia, the undisputed giant of microchips for artificial intelligence; the FTC that of Microsoft, where it has already been investigating the 'symbiosis' with OpenAi since January. Google and its links with another artificial intelligence firm, Anthropic, are also in its sights.

The latest move, which has seen formal requests to hand over documents relating to the past two years, was revealed by the Wall Street Journal. Microsoft, now at the top of the stock market cap ranking ahead of Nvidia, which is challenging the dethroned Apple for second place, hired Inflection co-founder Mustafa Suleyman, an artificial intelligence pioneer with a past at DeepMind and Google, and almost all of its employees in March. He agreed to pay $650 million to the company through a licensing deal to resell its technology and put Suleyman in charge of a brand new division, Microsoft AI. He also promised to repay investors in the start-up in the future from the proceeds of the business.

It was enough to worry the FTC. Companies are required to report any acquisition exceeding 119 million to the federal authorities for examination of the impact. The complex partnership mode, in short, in the eyes of the Ftc could conceal a significant merger. If tech brands often buy start-ups to absorb talented personnel, a formula christened 'acquihire', in this case Microsoft would have dribbled any risk of a federal stop altogether.

Ftc chairman Lina Khan is the inspirer of an aggressive overall government strategy to curb what she sees as the overwhelming power accumulated by Big Tech. Now she is transferring this approach to AI: the fear is that the cutting-edge frontier could end up dominated as never before by a very few giants who take over all the most promising start-ups, eliminating all competition and control when a particularly sensitive technology is at stake. San Francisco-based Inflection has developed one of the leading LLMs, the 'large language models' at the heart of AI, and an advanced chatbot christened Pi. Microsoft, moreover, has long had huge investment pacts with OpenAI and its ChatGPT. It is precisely OpenAI that has recently admitted that alarms about the use and non-use of AI, in politics and not only in business, have reason to exist: it has had to remove accounts of Russian, Chinese, Iranian and even its Israeli ally groups that were using its technology for propaganda campaigns fed by disinformation and fake users.