In a recent legal battle, tech company Singular Computing, based in Massachusetts, filed a lawsuit against Google, accusing the search giant of stealing its patented AI technology for Tensor Processing Units (TPUs). Singular claims that Google used the copied technologies as a basis for integrating AI features into services such as Search, Gmail, and Translate.

According to Singular, the alleged infringement occurred after meetings between Google and Singular’s founder, Joseph Bates, from 2010 to 2014. During these meetings, Bates shared his ideas, which Singular now claims were used without permission in the development of Google’s TPUs. Jurors in a Boston court were presented with emails from Google’s chief scientist, Jeff Dean, discussing the suitability of Singular’s technology for Google’s products.

Google’s defense, presented by lawyer Robert Van Nest, argued that the team behind the chipsets never met with Singular’s founder and independently developed the designs. Van Nest referred to Bates as a “disappointed inventor” and disclosed that Singular had previously approached other AI-focused companies, including Amazon, Meta, Microsoft, and OpenAI. He emphasized that Singular’s technology was prone to “incorrect” calculations due to its use of approximate math, asserting that Google’s chipsets were “fundamentally different” from those described in Singular’s patents.

Interestingly, pretrial documents revealed that Singular Computing initially sought up to $7 billion in damages, but during the trial, its lawyers were seeking $1.67 billion. The trial, which could extend for several weeks, delves into the alleged infringement related to the second and third versions of Google’s TPUs, introduced in 2017 and 2018, respectively.

This legal dispute adds to Google’s history of facing lawsuits, including the recent high-profile case with Epic Games. Google is also contending with a separate antitrust lawsuit related to the Play Store, in which it announced a $700 million settlement. As the trial unfolds, the technology community awaits the outcome and its potential impact on the ongoing development of AI technology.