With the advent of any new electronic technology, there is the potential for it to be misused. Napster was an emerging system for sharing audio files. While there could have been acceptable uses of Napster, it was swiftly determined by the courts that Napster was also being improperly used to share copywritten songs.
The potential for misuse now exists in the new world of artificial intelligence (AI). AI has a vast array of potential benefits. Regardless of the type of end use, AI is only as good or useful as the information that is fed into the training system. Therein lies one risk that protected information may be improperly used to educate an AI system.
Copyright protects original works of authorship that are fixed in a tangible medium of expression. 17 U.S.C. § 102. Under 17 U.S.C. § 106, the owner of a copyright has the exclusive right to do any of the following:
(1) to reproduce the copyrighted work in copies or phonorecords;
(2) to prepare derivative works based upon the copyrighted work;
(3) to distribute copies or phonorecords of the copyrighted work to the public by sale or other transfer of ownership, or by rental, lease, or lending;
(4) in the case of literary, musical, dramatic, and choreographic works, pantomimes, and motion pictures and other audiovisual works, to perform the copyrighted work publicly;
(5) in the case of literary, musical, dramatic, and choreographic works, pantomimes, and pictorial, graphic, or sculptural works, including the individual images of a motion picture or other audiovisual work, to display the copyrighted work publicly; and
(6) in the case of sound recordings, to perform the copyrighted work publicly by means of a digital audio transmission.
Thus, for example, AI should not be able to make or distribute copies of a copyrighted work. Furthermore, AI should not be able to prepare derivative works based on a copyrighted work.
Comedian, actress, and author Sarah Silverman, along with others, have alleged that OpenAI and Meta Platforms have directly and vicariously infringed their copyright rights in training their AI systems (e.g., ChatGPT and MetaAI). See Sarah Silverman et al. v. OpenAI Inc. et al., case no. 3:23-cv-03416, Northern District of California; and see also Kadrey et al. v. Meta Platforms, Inc., case no. 3:23-cv-03417, Northern District of California. For instance, the creators allege that OpenAI and Meta Platforms have improperly made copies of their protected works (e.g., books) and also improperly made derivative works based on those protected words. Conversely, the AI platforms have asserted that it is fair use of the works to train their AI systems, and that the generated output does not resemble the copyrighted works. Much like Napster, there also appears to be an effort to direct responsibility for improper use to the users of the AI systems.
There are also other potential concerns with generative AI. As one example, a radio host, Mark Walters, has alleged that ChatGPT defamed him with a fabricated and false report about fraud and embezzlement to a reporter. See Mark Walters v. OpenAI LLC, case no. 1:23-cv-03122, Northern District of Georgia. OpenAI is challenging the defamation claim on the basis that both Mark Walters and the reporter knew the report to be false.
In view of copyright infringement accusations, AI providers are increasingly providing indemnity protections for users of their AI platforms. On October 12, 2023, Google joined others in offering intellectual property indemnity with their generative AI (see https://cloud.google.com/blog/products/ai-machine-learning/protecting-customers-with-generative-ai-indemnification). Google also clarified that its general services indemnity covers infringement allegations pertaining to the training data used by Google for its generative AI.
While intellectual property rights and personal rights are certainly important issues, this evolving technology impacts a broad range of industries, and there are other important considerations such as regarding competition and bias. On October 30, 2023, President Joe Biden issued an executive order (https://www.whitehouse.gov/briefing-room/presidential-actions/2023/10/30/executive-order-on-the-safe-secure-and-trustworthy-development-and-use-of-artificial-intelligence/) regarding the standards for development of AI, which will be implemented through legislative action and federal agencies. In one aspect, the USPTO is tasked with publishing guidance on inventorship and the use of AI within 120 days of the executive order. Also, no later than 270 days after the executive order, the USPTO shall consult with the Director of the Copyright Office and provide recommendations to the President regarding the scope of protection for works produced using AI and the treatment of copyrighted works in AI training.
There is also an ongoing international effort to establish standards for the development of AI (e.g., https://www.gov.uk/government/news/countries-agree-to-safe-and-responsible-development-of-frontier-ai-in-landmark-bletchley-declaration).
It is a brave new world.
If you have any questions about AI infringement, contact an attorney to seek advice.