Microsoft CEO accuses Google of AI data grab

The trial against Google is the first major antitrust case brought by the US since it sued Microsoft in 1998.

You're missing out! Subscribe to the full version of DECODED Technology here.

Microsoft CEO, Satya Nadella, has accused Google of “locking up” the content that “feeds” developing large language models (LLMs) during his testimony at Google’s ongoing antitrust trial in Washington DC.

The landmark US trial against Google, is the first major antitrust case brought by the US since it sued Microsoft in 1998 and is viewed by many as an inflection point for Big Tech dominance.

Google already stands accused of monopolising the internet by unfairly dominating competition from other search engines by paying billions of dollars to Samsung and Apple to ensure Google’s place as the default web browser on many smartphones.

In his testimony yesterday (2 October), Nadella stated that Google had also divided up online content via these same default agreements with in expensive and exclusive deals with publishers.

Nadella’s statements were made in response to Judge Amit Mehta’s inquiry as to whether AI could level the playing field between Google and its competitors.

Microsoft has already invested more than $10bn into OpenAI’s LLM ChatGPT, whose release to the public back in November 2022 is often credited as the reason for AI’s subsequent coverage and investment boom.

However, Google’s billion dollar deals with content providers could curb the data available to train ChatGPT.

“When I am meeting with publishers now, they say Google’s going to write this check and its exclusive and you have to match it,” Nadella reportedly stated in court according to Reuters.

Exclusive access to daily industry insights. Select from over 20 different sectors, from Medical Devices to Technology, by subscribing today.

Speaking on Nadella’s comments, senior analyst at GlobalData Beatriz Valle agrees that Google “absolutely dominates” its competitors, however, does admit that Microsoft’s accusations feel potentially hypocritical.

“Nadella’s comments on training of LLMs on Google’s content are hardly surprising: since Microsoft added GenAI capabilities to Bing in February, the two companies have been locked in a battle for dominance of the search engine market” Valle explains.

Whilst Microsoft has invested copious amounts of dollars into ChatGPT to try and gain an upper hand on Google, Valle points out that Google has also invested money into creating the dominating LLMs of today.

“Market forces are at play here, and it is becoming apparent during this trial that Google’s search dominance is down to user choice as well as business practices,” she adds.

“Also, we are not comparing apples with apples here: Microsoft has promoted Bing as an AI-powered chatbot because, let’s be honest, as a search engine it cannot really compete against Google’s,” she continues.

Whilst Valle admits that Google leads over Microsoft’s search engine, she does state that Google has been “shy” in incorporating its Bard LLM into Google Search.

“Microsoft seems to hold the competitive advantage in this area,” she concludes, “because it controls OpenAI’s powerful LLM, so I really don’t get why Nadella complains about content to train LLMs as GPT-4 is known to be the most powerful LLM today.”

If you sign up to just one newsletter, make sure it's this one. Select your industry of choice to receive gold standard intelligence direct to your inbox.

By Alice Nunwick