US antitrust enforcer says 'urgent' review of AI control by Big Tech needed

Unlock Editor’s Roundup for free

The US’s top antitrust enforcer will look "urgently" at the artificial intelligence sector, following concerns that power over the transformative technology is being concentrated among a few deep-pocketed players.

Jonathan Kanter said in an interview with the Financial Times that he was looking at "monopoly choke points and the competitive landscape" in AI, including everything from computing power and data used to train large language models to cloud service providers , engineering talents and access. in essential hardware such as graphics processing unit chips.

Regulators are concerned that the nascent AI sector is "at the peak of competition, not the floor" and must act "with urgency" to ensure that the already dominant tech companies don’t control the market, Kanter said.

"Sometimes the most meaningful intervention is when the intervention is in real time," he added. "The beauty of that is you can be less invasive."

Kanter, now in his third year at the Justice Department, along with the Federal Trade Commission has led a tougher antitrust approach, suing tech groups like Google and Apple for what the US government claims are unfair monopolies in services, including app stores, search engines. and digital advertising. He has worked closely with FTC Chairwoman Lina Khan.

He said regulators were looking at the generative AI sector and examining the competitive landscape in microchips.

Kanter said the GPUs needed to train LLMs had become a "scarce resource." Nvidia dominates sales of cutting-edge GPUs and its market capitalization overtook that of Apple on Wednesday to become the world’s second most valuable listed company.

Kanter pointed to government initiatives to boost domestic manufacturing, including $39 billion in incentives in the Chip Act, but added that antitrust regulators were looking at how chipmakers decide to share their most advanced products amid demand unbridled.

"One of the things to think about is the conflict of interest, a big toe on the scale because they’re afraid to enable a competitor or they’re helping to support a client," Kanter said. "If decisions are made that show companies don’t care about maximizing profits or generating shareholder value, but more about looking at competitive consequences," then that would be a problem.

Since the sensational release of OpenAI’s ChatGPT chatbot in November 2022, an arms race has erupted as companies rush to secure multi-billion dollar partnerships with some of the most promising AI companies and those building models and applications based on the technology.

Emblematic of such deals is Microsoft’s $13 billion investment in OpenAI, which came with rights to the start-up’s intellectual property and a share of its profits, but fell short of a full buyout.

However, the FTC, as well as UK and EU competition watchdogs, have said they will investigate the relationship alongside Google and Amazon’s multibillion-dollar deals with rival Anthropic.

In March, Microsoft CEO Satya Nadella hired Mustafa Suleyman, the founder of another AI start-up called Inflection, and most of his 70-person staff to create a new consumer AI unit. Some industry watchers saw the deal as a tactic to circumvent antitrust laws and escape an official investigation.

"Acqui-angers are something that antitrust enforcers" will be looking at, Kanter said, while declining to comment on any specific transactions. "We are not using stylistic or formal features like these companies [explain these deals]. What we look at are market realities.

"We are focused on the facts. If the form is different but the substance is the same, then we will not hesitate to act," he added. "We look at what are the raw materials to produce a product. Whether it’s steel or engineers, it fits within the traditional paradigm of what we’re interested in."

Microsoft has pushed back against allegations that it exercises undue influence or de facto control through its investments and cloud computing services. It has also invested in France’s Mistral and put $1.5 billion into Abu Dhabi’s AI group G42.

"The partnerships we are pursuing have significantly increased competition in the market," technology group president Brad Smith told the FT. "I could argue that Microsoft’s partnership with OpenAI has created this new AI market," and without its help, the start-up "wouldn’t have been able to train or distribute its models."

Asked why Microsoft didn’t buy Inflection, he said: “We didn’t want to own the company. We wanted to hire some of the people who worked at the company.”

Deja una respuesta

Tu dirección de correo electrónico no será publicada. Los campos obligatorios están marcados con *

Translate »

0