How the Alliance Between OpenAI and News Publishers will Impact Journalism

Why Trust Techopedia
KEY TAKEAWAYS

  • An alliance is forming between OpenAI and major news publishers, with partnerships including News Corp, Vox Media and The Atlantic.
  • These deals allow OpenAI to train its language models on content from top publications, providing a new revenue stream for publishers.
  • Despite potential benefits, such partnerships may harm the impartiality and future of journalism by competing directly with traditional publications.
  • Financial pressures have driven publishers to seek these partnerships, potentially undermining the mobility of new journalists and the credibility of the content.
  • While AI is here to stay, its integration in newsrooms could devalue journalism and compromise ethical standards in the industry.

An unholy strategic alliance is forming between OpenAI and major news publishers, with the artificial intelligence giant announcing partnerships with Vox Media and The Atlantic, following a deal with News Corp in the last few weeks and with The Financial Times last month. 

These partnerships will enable the world’s biggest AI startup to train its large language models (LLMs) on articles and intellectual property produced by these brands, including publications such as The Wall Street Journal, The New York Post, Australia’s The Daily Telegraph, and The Sun in the UK. 

On first glance, these partnerships seem mutually beneficial. OpenAI gets to train its models on high-quality content produced by journalists, and publishers get access to an additional revenue stream in return.

Unfortunately, it’s hard to see these partnerships as being good for the impartiality or longevity of news publishing and, by extension, journalism. Or, as The New York Times’ lawsuit against OpenAI argues — the AI lab is producing a service that undermines the core business model of news publishing.

Why the OpenAI Deal is Bad for News Publishers in the Long Run

News publishing has been on the decline for quite some time, both in terms of ad revenue and readership.

Aaccording to a Pew Research report released in November 2023, the average number of unique visitors to the websites of the U.S’s top 50 newspapers declined 20% to under nine million in the fourth quarter of 2022.

Given these financial pressures, it’s unsurprising that the publishers outlined above have opted to partner with OpenAI to secure an additional revenue stream before the competition gets a chance to.

Andrew Gamino-Cheong, CTO and co-founder at Trustible, told Techopedia:

“When I saw the deluge of recent announcements, my first thought was: if you can’t beat em, join em.

 

“There will only be so many news sources supported by the big AI players, and better to be one of those sources than lose out entirely.”

The problem is that no matter how much OpenAI pays these publishers, they’re creating a service that directly competes with traditional publications. Models trained on content taken from News Corp properties, Vox Media, and The Atlantic, from everything we can ascertain, will literally be used to create AI-generated news content that takes clicks away from human journalists.

While AI can’t replace the on the ground reporting and expertise of an experienced journalist, it can oversaturate the market and make journalism less profitable than it already is. This won’t be as damaging to established journalists, but it has the potential to weaken mobility of those who are just starting out.

At the same time, it’s hard not to see these publishers as betraying the writers that helped to build their brands. After all, it is unlikely that all of the authors featured on these sites consented to having their work scraped by AI models.

How AI-News Publisher Deals Impact Credibility

When examining these partnerships we have to consider that they are taking place alongside the entry of AI into the newsroom. For instance, OpenAI announced yesterday that it would be collaborating with WAN-IFRA on a global accelerator program to assist over 100 news publishers “in exploring and integrating AI in their newsrooms.”

The program will last three months per team and includes learning modules, hands-on workshops, a mini hackathon, and a showcase. Most notably, “participants will leave the program with a clear idea and plan on how to roll out AI in their newsroom.”

These types of partnerships have the potential to damage the credibility of the publication — it becomes unclear how much work is actually produced by human journalists while giving an unaccountable and potentially biased third party blackbox model a role in content creation.

The use of AI in these newsrooms could be anything from the occasional query to generating headlines and drafting questions or outputting complete artifles. The more AI is involved in the process, the greater the potential for bias.

Things also get trickier if a newsroom uses AI and has a commercial relationship with a vendor providing it.

For example, The Atlantic says that it’s building a site called Atlantic Labs “to figure out how AI can help in the development of new products and features to better serve its journalism and readers — and will pilot OpenAI’s and other emerging tech in this work.”

While The Atlantic is a fine publication, and newspapers are generally adept at both looking in the mirror and keeping walls up between a free press and a publication’s relationships with vendors,it may not aid debate around the ethical use of AI.

The Bottom Line

Make no mistake, generative AI is here to stay, but publishers should be extremely cautious about using it in news production.

Licensing out content to a competing service may be a short-term measure that does little for the longevity of the industry and in the worst case, may devalue journalism as a profession.

Related Terms

Related Article