Could we see law firms questioning Google and OpenAI on the use of law blogs published by law firms and other legal content in these AI engines?
The issue is being raised by larger publishers, as reported by the Wall Street Journal on Sunday.
The emerging awareness has set up a war between the forces behind the inputs and the outputs of these new artificial intelligence tools, over whether and how content originators should be compensated. The disputes threaten to throw sand into the gears of the AI boom just as it seems poised to revolutionize the global economy.
Artificial intelligence companies including OpenAI, its backer Microsoft, and Google built generative AI systems such as ChatGPT by scraping oceans of information from the internet and feeding it into training algorithms that teach the systems to imitate human speech. The companies generally say their data use without compensation is permitted, but they have left the door open to discussing the issue with content creators.
OpenAI and Google say they train their AI models on “publicly available” information.
Their claims get broader when using Fair Use to say they have the right to use the content so long as it’s not in its same form.
Tech companies have pointed to the legal doctrine of fair use, which permits the use of copyrighted material without permission under some circumstances, including if the end product is sufficiently different from the original work. AI proponents say free access to information is vital for technology that learns similarly to people and that has huge potential upsides for how we work and live..
Interesting road ahead for law blogs, journals, and reviews. Primary law – codes, cases, regulations, and constitutions – are already in the AI engines. Secondary law – insight and commentary – is needed to breathe life into this law.
Will law firms sue AI companies claiming a copyright violation? It’s possible, but unlikely. Fighting a huge company in an evolving area of the law, with other firms unlikely to join in, would be difficult or impossible.