Google executives downplayed the company’s position on artificial intelligence (AI) in recent testimony during its federal antitrust trial, saying the company has tried to be slow and cautious when implementing the technology because of its “dangerous” power. 

The Department of Justice (DOJ) didn’t see it that way during the trial. It said Google became more advanced in GAI and chose not to release the technology sooner for fear of losing its monopoly in search, according to Bloomberg.

It wasn’t until Microsoft revealed it had an AI chip, code name Athena, that could train large language models that Google showed its hand.

The company has touted the use of AI for years. Just not publicly. Building sophisticated models just doesn’t happen overnight.

Alphabet CEO Sundar Pichai last week during an earnings call said “we are seeing strong demand for the more than 80 models, including third-party and popular open source in our Vertex, search, and conversational AI platforms, with the number of customers growing more than 15x from April to June.”

He said more than 70% of GAI unicorns are Google Cloud customers, including Cohere, Jasper, and Typeface.

“We provide the widest choice of AI supercomputer options with Google TPUs and advanced NVIDIA GPUs, and recently launched new A3 AI supercomputers powered by NVIDIA’s H100,” Pichai said during the earnings call.

During the trial, per Bloomberg, Pichai said that Google’s multibillion-dollar search deals simply give consumers the very best of the internet and the very best of the products that partners sell.

The DOJ argued that when Microsoft’s deal with OpenAI to integrate ChatGPT into Bing became public, Google mandated, internally, the integration of AI into its products.

Google said for years it declined to use AI in its search engine. That’s until competition rose and the company needed to put more space between it and others. Pandu Nayak, Google’s top search quality executive, said the company wanted people to understand its ranking systems.

That changed in 2015, when Google decided to begin integrating new machine learning into its technologies, he said. 

The company has since integrated several similar algorithms into its search engine to help it better understand the context of user queries. The longer and more complex the query, the better understanding of the intent, Google claims.

Those algorithms rely on much less user data, sometimes not even needing search data at all, Nayak said, but the query can still identify trends in user behavior. Google has always said it keeps strict control over how they are used in search.

The trial was expected to last 10 week from the September 12, start date, which means it should last another four weeks. The judge is trying to determine whether Google closed off avenues for other companies trying to compete.

Based on case filing, Google has paid billions to cellular device manufacturers and browser developers to ensure it remains the default general search engine for mobile phones, laptops, and computers. Google’s contract with Apple is said to cost it about $26.3 billion, which also ensures its partners do not work with competitors. 

Google execs downplayed the company’s position on AI during its federal antitrust trial, saying the company has tried to be slow and cautious when implementing the technology because of its “dangerous” power.