OpenAI’s ChatGPT May Be Rejecting Prompts That Violate Copyright

It appears the AI software is delivering error notices when asked to work with content that is copyrighted

chatbot Dall-E
(Dall-E)

Try to ask ChatGPT to produce content that is copyrighted and you may get a “no.” The AI-generated software appears to have been programmed to do so.

The disruptive technology backed by Silicon Valley titans that debuted in November may have been recently programmed to identify and prevent intellectual property theft, which many in Hollywood and other industries may see as a reprieve.

Representatives for OpenAI didn’t immediately respond to TheWrap’s request for comment.

The tweaks appear to be related to OpenAI, the parent company of ChatGPT, announcing it would soon launch paid version of the software, ChatGPT Professional. TechCrunch reported that news, citing the company’s statement on its official Discord server. OpenAI is “starting to think about how to monetize ChatGPT” to “ensure [the tool’s] long-term viability,” the statement said.

There appears to be a desire for a professional version of the software. On ChatGPT official LinkedIn page, the company posted a poll asking “Are you planning to use ChatGPT in your work this year?” and with two weeks still left on the poll, more than 7,300 voters have responded. More than 20% answered “not sure,” while 72% said “yes,” and 8% said “no.”

Tangentially, the copyright issue is surfacing in K12 education — just as ChatGPT is being digested by Hollywood executives, actors and entrepreneurs.

Schools in New York City and the Washington, D.C.-Maryland-Virginia (DMV) area and elsewhere have blocked students from accessing the artificial intelligence tool, concerned attempts to use the AI-generated responses to write academic papers constitutes plagiarism. While students can access the software on their personal mobile devices and computers, the tool is blocked on their school-issued devices.

Comments