Not even the Associated Press, a journalistic standard-bearer, can avoid the OpenAI wave.
The esteemed wire service announced Thursday that it has entered a partnership with the creator of generative AI platform ChatGPT to license its text archive from 1985 onward to help it develop its product’s knowledge base. In exchange, OpenAI will share its technology with the AP to see what role it can play in its news products, which do not currently use generative AI.
The partnership, which was first reported by Axios, comes while the ChatGPT creator is under federal investigation.
ADVERTISEMENT
“Generative AI is a fast-moving space with tremendous implications for the news industry. We are pleased that OpenAI recognizes that fact-based, nonpartisan news content is essential to this evolving technology, and that they respect the value of our intellectual property,” Kristin Heitmann, a senior vice president at the AP and its chief revenue officer, said in a statement.
The embrace of a generative AI platform is a first for the AP, though it has previously planted roots in the artificial-intelligence space. The company has partnered with various AI start-ups to develop AI bots, some of which have reported company business earnings and local sports games over the last decade. The AP said in its statement on Thursday that it has also used AI software to aid in audio and video transcription from news events.
An AP spokesperson would not say whether the new deal would impact those previous partnerships.
Earlier this year, the AP announced that it would develop five AI projects with local newsrooms in Texas, Minnesota, Michigan, Pennsylvania, and Puerto Rico, some of which may employ OpenAI’s GPT-3 technology. An AP spokesperson clarified that Thursday’s deal was separate from its local news AI initiative.
It was unclear what editorial staffers or departments would be involved in the OpenAI deal or how it would report on the company moving forward. An AP spokesperson would not clarify what staffers would be involved in the initiative, instead pointing to a news story with an “Associated Press” byline discussing the deal.
The wire service’s del with OpenAI comes as newsrooms ranging from The New York Times to Insider to G/O Media-owned shops figure out how to incorporate generative AI into their product after ChatGPT’s whirlwind debut last year—all without sacrificing the human journalists who deliver the news. Some outlets, such as the Times, have taken cautious approaches to researching AI’s capabilities, while G/O Media’s decision to test-publish stories written by AI has rankled staffers within Gizmodo.
The announcement also coincides with OpenAI facing intense scrutiny from legislators and government regulators still wary of its capabilities.
The Washington Post reported Thursday that the Federal Trade Commission was investigating OpenAI to see whether it has “engaged in unfair or deceptive privacy or data security practices” or whether it has acted in ways that could harm its users, “including repetitional harm.”
An AP spokesperson would not comment on whether the the company was aware of the investigation or whether it held any concerns regarding data security, privacy, or risks to consumers in partnering with OpenAI. A press line for OpenAI did not respond to an immediate request for comment.
