ChatGPT developer OpenAI reaches deal with Associated Press giving access to news archive back to 1985 amidst data controversy
ChatGPT developer OpenAI and Associated Press have reached a deal that will give the AI company access to AP’s archive of news stories dating back to 1985, as US regulators step up their scrutiny of AI companies’ use of data.
The deal will see “OpenAI licensing part of AP’s text archive, while AP will leverage OpenAI’s technology and product expertise,” the organisations said.
They did not disclose financial details.
The move comes after the US Federal Trade Commission launched a probe into OpenAI last week that includes examining its use of third-party data to train its AI tools.
AI companies have recently faced action from social media services such as Twitter, book authors and others alleging that the use of their data for AIs infringes their rights.
Generative AI services such as ChatGPT ingest large amounts of material from varied sources and use that material as the basis for “new” generated content.
However, observers have frequently discovered barely-disguised material from original sources within the generated content.
In January, for instance, AI-generated articles published by online media firm CNET were found to contain extensive structural and phrasing similarities to articles taken from human-written articles from competitors such as Forbes – in addition to a number of factual errors.
CNET paused its use of AI-written articles at the time, and in June published an AI policy pledging transparency.
AI-generated articles are currently still available on CNET sister site Bankrate, having been corrected for errors found by third-party journalists.
Such gaffes have raised questions over how AI companies will gain access to training data in the future, and the AP deal is an example of how such firms may be able to guarantee legal access to the needed material.
Questions also remain about how media firms may use AI, with AP saying it would use the partnership with OpenAI to help it understand responsible use cases and potentially make use of generative AI in future products and services.
AP has been a leader in the use of automation in producing routine articles such as corporate earnings reports and local sporting events.
The agency also introduced an AI-powered search engine for partners earlier this year that can access articles and photos using decriptive language.