The company actually filed a similar lawsuit against OpenAI, the maker of ChatGPT, last month. (OpenAI did not immediately respond to a request for comment on the lawsuit.)
The complaint alleges that Google has been “secretly stealing everything created and shared on the Internet by hundreds of millions of Americans” and using that data to train its AI products, such as chatbot Bard.
The complaint also notes that Google has taken “nearly our entire digital footprint,” including “creative and written work” to build its AI products.
Representatives for Google, Alphabet and DeepMind did not immediately respond to a request for comment.
The complaint refers to a recent update to Google’s privacy policy that explicitly states that the company may use publicly available information to train its models and AI tools like Bard.
In response to the Verge’s previous report on the update, the company said that its policy “has long been transparent that Google uses publicly available information on the open web to train language models for services like Google Translate. This most recent update simply shows that newer services like Bard are also included.”
The process comes as a new set of AI tools has garnered widespread attention in recent months for its ability to create text and images in response to user requests.
The large language models that support this new technology are able to do this by training on vast online databases.
In the process, however, the companies are also attracting increased legal scrutiny over copyright issues of the works scanned into these datasets, as well as their apparent use of personal and potentially sensitive data of regular users, including children’s data, according to Google’s lawsuit.
Tim Giordano, one of Clarkson’s attorneys who brought the lawsuit against Google, said in an interview with Google CNN.
“Our personal information and data is our property and value, and no one simply has the right to take it and use it for any purpose,” he said.
The lawsuit seeks an injunction in the form of a temporary freeze on commercial access and commercial development to Google’s generative AI tools such as Bard.
It is also seeking unspecified damages and payments as monetary compensation to persons whose data Google claims misappropriated. The company says it has collected eight plaintiffs, including a minor.
Giordano contrasted the alleged benefits and harms of how Google typically indexes online data to support its main search engine with new claims that it collects data to train AI tools.
Through its search engine, he said, Google can “provide a link attributed to your business that might lead someone to buy it or do business with it.”
Giordano added that data mining to train AI tools creates “an alternate version of the job that fundamentally changes the incentives for anyone who needs to buy the job.”
While some internet users are accustomed to having their digital data collected and used for search results or targeted ads, this may not be the case for training AI.
“People can’t imagine that their information will be used in this way,” Giordano said.
Ryan Clarkson, a partner at the law firm, said Google needs to “create an opportunity for people to opt out” of their data to train AI while retaining the ability to use the internet for their daily needs.
This content was originally published Google sued for stealing user data to train its AI tools On site CNN Brazil.
source: CNN Brazil
“Friendly zombie fanatic. Analyst. Coffee buff. Professional music specialist. Communicator.”
