Engine submitted comments to the U.S. Patent and Trademark Office last week in response to the agency’s request for public comment on “the impact of artificial intelligence (“AI”) technologies on intellectual property law and policy.”
In order to develop artificial intelligence technologies, startups and other companies need to input vast amounts of data into AI systems. The USPTO is currently reviewing how copyright law applies (or should apply) to the process of AI ingesting content and using it as the AI system learns its functions. Under existing law, it is lawful to ingest content and use it for training, tuning, or testing AI systems. Such uses either do not invoke copyright law (because the uses involve uncopyrightable data or noninfringing uses) or are permissible fair uses of material.
In our view, the relevant legal framework is working well. We are concerned that changing the law to increase the copyright burden on startups and AI companies would harm innovation, particularly for startups that need to pull-in outside content to build their AI datasets. As we say in our comments:
“AI development is thriving, and the recent explosion in promising AI technology occurred under the current copyright framework. Before making any changes to this framework, policymakers should carefully consider how any changes would impact the current trajectory of ubiquitous, varied, and exciting AI technologies. And policymakers should avoid making any changes that might hamper innovation.”
You can read our full comments here.
