Gen AI raises copyright issues. Could blockchain solve them?

As media bigwigs ward off the threat of gen AI, the spotlight might shine on blockchain as a viable solution.

In the news

The media vs. generative AI wars are raising the classic question: whether to “beat ’em” or “join ’em.” And as the various media players make their moves, another question is emerging: whether to ward off the threats of one powerful technology—generative AI—with another one: blockchain.

At issue is generative AI makers’ reliance on online content to train their models and supply answers to user queries—sometimes with word-for-word outtakes from published articles.

The New York Times has taken a decidedly “beat ‘em” approach by filing a lawsuit against OpenAI and Microsoft, claiming that by using its content to train their generative AI programs, the tech titans are threatening its livelihood.

Meanwhile, global media publisher Axel Springer joined forces with OpenAI by signing a blockbuster deal to make its Politico, Business Insider, Bild and other content available for model training and ChatGPT responses. In exchange, OpenAI will clearly brand the source of the material and link to the original source.

For its part, US media giant Fox Corp. is leveraging its previous investment in blockchain to unveil a platform that can track where and how its content is used online. In addition to using the platform, dubbed Verify, itself, Fox also aims to license it to other industry players—suggesting that “the tool will give them leverage in their negotiations with AI companies.”

All three of these cases highlight the role of media-developed content in a landscape increasingly influenced by generative AI. But the Fox case is interesting in another way, too: as a real-world example of blockchain’s emerging utility. Blockchain is best known as the engine on which cryptocurrency runs, but use cases are taking hold in multiple industries.

In healthcare, blockchain is being used to enhance information sharing and transparency, notably around medical records and pharmaceutical supply chains. In banking and financial services, the technology is used to manage payments and loans and administer smart contracts by such major firms as Bank of America and JP Morgan Chase. And in agriculture, blockchain has quickly become an essential tool for food supply chain traceability. 

For part of the general public, blockchain may be widely associated with crypto-catastrophes such as the 2023 downfall of FTX CEO Sam Bankman-Fried. But in business, the story is quite different. And with the mainstreaming of generative AI, blockchain may be a very good complementary partner.

The Cognizant take

In the public mind, there’s only room for one It Girl technology at a time; these days, that means folks choose among virtual reality, generative AI, and blockchain, then root for their fave as if it were a sports team.

The reality, says Duncan Roberts, a Senior Manager at Cognizant Research, is far different—and far more productive. “Gen AI and blockchain are actually quite complementary technologies,” he notes. For instance, generative AI tends to be a highly centralizing tech. “Very few have the ability to run their own large language model, so you’re beholden to a central service provider, and thus to what they decide their LLM will do,” Roberts says. Blockchain, on the other hand, is all about decentralization—taking trust away from a single authority and embedding it into the mechanisms of the blockchain itself.

It’s a topic Roberts has written about; he advises thinking of blockchain as a sort of auditor “that complements gen AI—and AI has issues around copyright tracking.” Regarding the recent goings-on at media companies, he adds that “Intellectual property management is a strong blockchain use case. Who created a given bit of content? Who should be rewarded for its use? Blockchain is useful anytime you need verification of a database.”

History shows that demanding the cessation of use of any given technology is seldom fruitful; those who take the “join ’em” approach—and can then commercialize the tech mechanisms supporting it—may be more likely to succeed.