Google, a major tech company listed on the NASDAQ as GOOGL, is currently attempting to have a class-action lawsuit thrown out. This lawsuit alleges that Google has infringed upon the privacy and property rights of “hundreds of millions of Americans” by collecting data for the purpose of training its artificial intelligence (AI) systems, including the generative AI chatbot known as ‘Bard.’
In the motion filed with the U.S. District Court for the Northern District of California on Tuesday, Google argued that the claims made against the company are founded on an erroneous premise. The premise in question is the assertion that training Generative AI models using information publicly available on the internet constitutes “stealing.”
Google’s counterargument was framed as follows: “Utilizing publicly available information for the purpose of learning does not constitute theft, nor does it qualify as an intrusion into privacy, conversion, negligence, unfair competition, or copyright infringement.”
Additionally, the company expressed concerns that allowing the lawsuit to proceed would have broader implications, stating that it could potentially undermine not only Google’s services but also the fundamental concept of generative AI.
The class action lawsuit, initiated in July by eight individuals who asserted they were representing “millions of class members,” made allegations against Google, including claims of privacy invasion, negligence, and copyright infringement.
The suit alleged, “Google has collected all our personal and professional data, our creative and copyrighted works, our images, and even our emails – virtually the entirety of our digital footprint – and is using it to develop commercial Artificial Intelligence (‘AI’) products. For years, Google gathered this data in secrecy, without providing notice or obtaining consent from anyone.”
The class action contended that the plaintiffs’ privacy and property rights were violated due to a change in Google’s privacy policy, which occurred a week before the lawsuit was filed, permitting data scraping for AI training purposes.
The plaintiffs argued that Google had “silently updated its online privacy policy recently to reinforce its stance that everything on the internet is fair game.” They further asserted that the term “publicly available” had never meant that such data was free to be used for any purpose.
In its response on October 17, Google maintained that the complaint consisted of “irrelevant actions by third parties and alarmist predictions regarding AI.” Additionally, the tech giant argued that the allegations in the complaint were unclear and did not adequately explain the primary accusations:
“Very little information sheds light on the fundamental issues, such as specifying what personal information of the plaintiffs was allegedly collected by Google, in what manner (if at all) that personal information is reflected in the output of Google’s Generative AI services, and how (if at all) the plaintiffs have suffered harm.”
This case is just one of several ongoing legal disputes in U.S. courts related to artificial intelligence.
For instance, in January, three visual artists filed lawsuits against various generative AI platforms, alleging that these companies used their artworks without permission or compensation to create the training datasets for their AI algorithms. In February, Getty took legal action against Stability AI, accusing the Stable Diffusion text-to-image AI program of unauthorized use of its images, thereby infringing on copyright and trademark rights associated with its watermarked photograph collection. In July, several authors initiated a lawsuit against OpenAI, claiming that their copyrights were violated during the training of the company’s AI systems.
Google has faced prior legal challenges related to improper data usage, and these previous cases could potentially impact the outcome of its current motion to have the lawsuit dismissed.
Beneficial legal precedent for Google
In 2005, Google effectively defended itself against a lawsuit brought by the U.S. Authors Guild. This legal action alleged that Google had committed “extensive copyright infringement” by making copies of published works for its search engine and archive.
In 2013, the court ruled in favor of Google, granting the company’s request for a summary judgment, leading to the dismissal of the claims made against it.
The U.S. Authors Guild appealed this decision, but in 2015, the court upheld the ruling, stating that Google’s “unauthorized digitization of copyright-protected works, the development of search functionality, and the display of snippets from these works constitute non-infringing fair uses” due to their “highly transformative” purpose.
Although the specifics of this case differ from the recent AI-related class action against Google, it does establish a legal precedent where the tech company successfully defended itself against allegations of copyright infringement related to its data collection and usage practices.
This might have been a consideration for the company when, on Tuesday, Google made an argument akin to the one endorsed in the Authors Guild decision. Google asserted that “simply providing information about a copyrighted book or on the same subject matter does not amount to infringement.” The usage is distinct or ‘transformative,’ and thus could potentially qualify as ‘fair use’ under U.S. copyright law.
Another illustration of this argument pertains to its generative AI chatbot, Bard, with Google contending that “the complaint must demonstrate a substantial likeness between Bard or its output and the copyrighted content in the book, but it does not make an effort to establish such similarity.”
The ultimate decision in this case is yet to be determined, but the tangentially relevant precedent appears to be in favor of Google. If the outcome of this case also aligns with Google’s position, it could have significant ramifications for several AI-related cases currently progressing through the U.S. legal system.
No Comment