Experts Dismiss OpenAI Patent Pledge as Vague Public Relations Stunt
OpenAI publishes a promise to only use its patents defensively, but intellectual property experts call the statement ambiguous and legally unenforceable. Critics argue the vague language allows the company to target competitors while protecting a surprisingly small patent portfolio.
OpenAI quietly publishes a pledge on its website promising to only use its patents defensively to support innovation. The company states that it will refrain from offensive patent actions as long as other parties do not threaten OpenAI, assert claims against it, or engage in activities that harm the company or its users. However, intellectual property experts quickly point out that this commitment is highly ambiguous compared to historically specific pledges like IBM's 2005 patent promise.
Legal experts criticize the vague wording of the document, noting that the exception for activities that "harm" OpenAI is broad enough to include legitimate competitors and fair critics of ChatGPT. Furthermore, this pledge covers very little actual intellectual property because OpenAI currently holds only a handful of granted patents. The company relies primarily on trade secrets, such as confidential training data and methods, rather than a robust patent portfolio.
The statement is entirely voluntary and lacks the legal weight of a binding contract or warranty. Syracuse University professor Shubha Ghosh emphasizes that this pledge does not prevent OpenAI from enforcing its copyrights, trade secrets, or terms of service agreements. Ultimately, experts view the announcement as mere public relations virtue-signaling designed to curry favor with regulators and the tech community without requiring any meaningful sacrifices.