AI: Is Microsoft and Nvidia's dominance damaging?
June 18, 2024
Microsoft and Nvidia have quickly taken prime positions in the artificial intelligence revolution. The tech giants have a combined market value of $6.6 trillion (€6.16 trillion), making them two of the three largest companies in the world.
Microsoft's recent success has been underpinned by its $13 billion bet on OpenAI, the startup behind the ChatGPT chatbot, while Nvidia can boast the world's most advanced chips that are vital to running high-end AI systems.
But their good fortune means competition authorities in the United States are circling. Earlier this month, the US Department of Justice (DOJ) and the Federal Trade Commission (FTC) reached a deal on how to investigate the two firms' dominant position in the AI space.
The FTC is set to focus on the close relationship between Microsoft and Open AI, whose parent company is a non-profit. The DOJ will, meanwhile, lead the investigation into Nvidia's competitive edge. The chipmaker has around 80% of the AI semiconductor market and is now valued at $3.32 trillion, up from just $364 billion two years ago.
"Big Tech has gained too much power in the last 15 years or so, and regulators have been asleep at the wheel," Simonetta Vezzoso, a lawyer and economist from Italy's Trento University, told DW. "Now they worry that they might be replaying the same game with AI and want to avoid that."
Smaller players depend on Big Tech
Startups need large amounts of data, storage and chip capacity to be able to train their AI chatbots, which is where regulators believe the tech giants hold too much power. There is some evidence that smaller players are being forced into exclusive, opaque deals to run technology from Nvidia, Microsoft and their rivals, which can give the already dominant players an even bigger advantage.
"The competition authorities want to protect innovation coming from startups. These deals come with a lot of strings attached, so Big Tech could be hampering that competition," Vezzoso added.
DOJ antitrust chief Jonathan Kanter told an AI conference at Stanford University in California earlier this month that "powerful network effects may enable dominant firms to control these new [AI] markets."
Recent mergers likely to be probed
Microsoft's $650 million acquisition of Inflection AI, the startup behind the Pi personal assistant app, in March also raised eyebrows as the deal may have been designed to avoid merger disclosure rules.
"Microsoft bought Inflection without buying it," Pedro Domingos, professor emeritus of computer science and engineering at the University of Washington, told DW. "They broke it into pieces, hired most of their staffers and paid off the investors."
Some policymakers believe the lack of scrutiny over earlier mergers saw Big Tech acquire hundreds of startups that could have gone much further to disrupt the tech sector, so their probes will also focus on the impact on innovation.
Vezzoso said that to remedy their past mistakes, antitrust regulators are keen to act more quickly and "reverse the burden of proof" from themselves to the tech giants. If necessary, she called for "very resolute measures on Big Tech."
"I would like to see them [regulators] be very assertive. If a big tech firm wants to buy a small startup, they should have to show that there is no anti-competitive issue," said Vezzoso, who is also an external consultant for the human rights group Article 19.
Impossible to read the future?
Domingos, on the other hand, thinks it's "comical" to launch antitrust lawsuits "not about harms that have been done, but about those that might happen in the future."
The author of the book "The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake Our World" noted how Meta CEO Mark Zuckerberg has repeatedly said that Instagram would not be the success it is today if Facebook hadn't bought it.
"Facebook gave [Instagram] a huge amount of infrastructure and expertise that they didn't have. If you fast forward, you can apply the same reasoning to Microsoft and Nvidia and the AI startups that they may be buying," Domingos added.
US regulators team up
US President Joe Biden vowed again this year to make scrutiny of Big Tech a priority for his administration and some legal experts have spotted a more collaborative approach from the FTC and the DOJ in cracking down on Silicon Valley's business practices.
"It used to be that the agencies divided the cases according to industry, but with this market being so large and important for antitrust enforcement, they are sharing responsibility and working hand-in-glove," Rebecca Haw Allensworth, a professor at Vanderbilt Law School, told The Guardian recently.
As the US presidential election approaches, just a short window of opportunity may remain for the Biden administration to take action, which could still be undone if Donald Trump wins the White House in November.
Tech giants face growing resentment
Domingos, meanwhile, noted how federal and state lawmakers have introduced nearly a thousand laws to regulate AI since ChatGPT was first released, noting that some policymakers have "a lot of hostility towards the big tech companies and want to use AI as a tool to attack them with."
The additional scrutiny is already having a chilling effect on the tech sector, where the big tech giants are growing increasingly scared of acquiring promising startups.
Merger and acquisitions hit a multiyear low last year in terms of deal value, falling below $300 billion, according to 451 Research. In 2022, the value of all buyouts was nearly $800 billion. The large strategic tech players such as Meta, Salesforce, Alphabet, Apple and Amazon made just 4 acquisitions last year, compared to 18 a year earlier, data from Capital IQ Pro shows.
"The big tech companies are now afraid to do acquisitions, which is damaging to the ecosystem because for a lot of startups — their fate in life is to be acquired — and everybody benefits," Domingos told DW.
Edited by: Ashutosh Pandey