January 8, 2024


Can OpenAI and news publishers co-exist?

It’s 2024 and in some ways, it’s a lot like 2023 again – we’re talking about artificial intelligence (AI). A lot has been said and written about AI in the past year. 2023 was a banner year for AI and seemed to occupy more people’s time and thoughts. Actors and writers went on strike for months in Hollywood, in part, demanding control over AI in storytelling, the OpenAI-Sam Altman saga, to name a couple.

As we enter 2024, AI will likely continue to dominate the news cycle. We didn’t have to wait too long as the New York Times sued OpenAI and Microsoft for copyright infringement in a landmark suit late last month. It’s likely to create a new battleground over the unauthorised use of published work to train AI models. This suit could set an important precedent and begs the question – can news publishers, OpenAI, and other AI models co-exist?


Almost a year ago, Microsoft unveiled “the new Bing” by introducing new AI features and models into its search engine. CEO Satya Nadella called it a new paradigm. Same with Google and its Project Bard. Both have stated that a new era of AI-assisted search is upon us. The one that arguably started it all was OpenAI and its ChatGPT.

Chatbots like ChatGPT function based on the model being fed tons of data over time. When you ask a question, it goes through its database and provides an answer. Whether it’s accurate is still up for debate, and there’s plenty of evidence to prove it’s often not. 

Let’s say you have a question about Covid-19 and ask the chatbot. It’ll provide information sourced from news articles. Does the chatbot have the authority to provide large swaths of information from a source? The News Media Alliance, representing over 2,000 publishers, said no. It stated AI companies regularly used information without authorisation and violated intellectual property laws.

The Alliance said AI companies regularly use content from media companies to train their models. It stated the models also copy and publish content from news sites as outputs to users. They argued it diminishes the high-quality human-generated content and harms publishers and AI models. Given these AI models are far from foolproof, these practices could harm AI companies since trustworthy information will become more difficult to find.

AI companies and their products have found themselves at the receiving end of lawsuits from news companies and publishers. Thomson Reuters accused a legal AI company of copying its content. Authors and writers have also spoken up, filed cases, and written letters to AI companies to protect writers and their content.

Coming to the New York Times case, the newspaper alleged that millions of its articles were used to train automated chatbots that now compete with the news outlet as a source of information. While not specifying any monetary value, it said the companies should be held responsible for “billions of dollars in statutory and actual damages”.

2024 will likely be another eventful year for AI companies. While attracting billions of dollars in investment, there’s also likely to be regulatory and legal attention. Lawsuits and potential regulation could probably slow things down for AI that leapt forward in 2023. For the news business, the Times lawsuit will set the stage for what sort of relationship they’ll have with AI companies. Can both co-exist?

VIEW: There’s room for both

There’s no doubt that the relationship between AI and the news industry is fraught right now. That doesn’t have to be the case in the future. Apple has been speaking to news publishers seeking permission to use their content in developing its generative artificial intelligence systems. The crux of the matter has to do with the news business, well, staying in business. That means being compensated for others using their work. Apple has offered multiyear deals worth at least $50 million.

The same goes for OpenAI. Last month, it struck a landmark deal with Axel Springer to “strengthen independent journalism in the age of artificial intelligence (AI).” It allows for Axel Springer content to be used in ChatGPT. Specifically, users will get summaries of selected news content from Springer’s companies like POLITICO, Business Insider, etc. This deal certainly sets a precedent for the relationship between AI companies and news publishers in the year(s) ahead.

OpenAI-Springer deal is meant to ensure transparency. In the answers and summaries that users get to their queries, there’ll be attribution and links provided to the full articles. So, the media company will get a certain amount of web traffic back to their site. For the news companies, this does seem like a lucrative opportunity in the face of a more uncertain ad revenue landscape. This also ensures that their content isn’t taken without permission.

COUNTERVIEW: Tough to see eye-to-eye

If a media company copied articles from another news outlet and posted them as its own, it would be a cut-and-dry case of copyright infringement. What if a tech company did the same to train its AI chatbot? It’s where things get murky. While tech companies might offer the “fair-use defence” in the USA, the legal and regulatory landscape across different countries isn’t nearly robust. These models can basically reproduce exact copies of articles.

The tricky part about all this is publishers now have to compete with the likes of OpenAI and Google. Take the latter as an example. Google’s Search Generative Experience (SGE) uses AI to create summaries for some search questions. If a publisher doesn’t want their content to be used for this, they need to use the same tool that would prevent them from appearing in Google search results. Also, publishers compete with Google for online ad revenue and depend on it for search traffic. See, it gets tricky.

What’s the guarantee that people will click on the link if the summary suffices? That’s the worry publishers have with OpenAI too. At the very least, traffic to their news sites will decrease to some extent. It once again comes down to money. How much is news content worth today? How much is the work done by journalists, often for weeks and months, worth? According to some reports, for OpenAI, it’s between $1 and $5 million a year. Does a news publisher think this is enough? Who gets to decide that amount?

Reference Links:

  • Inside the News Industry’s Uneasy Negotiations With OpenAI – New York Times
  • OpenAI in talks with dozens of publishers to license content – Moneycontrol
  • ChatGPT and other AI chatbots rely heavily on copyrighted news media, say publishers – Fast Company
  • The New York Times’s OpenAI lawsuit could put a damper on AI’s 2024 ambitions – Fast Company
  • Apple Explores A.I. Deals With News Publishers – New York Times
  • Partnership with Axel Springer to deepen beneficial use of AI in journalism – OpenAI
  • OpenAI’s news publisher deals reportedly top out at $5 million a year – The Verge
  • AI’s future could hinge on one thorny legal question – The Washington Post

What is your opinion on this?
(Only subscribers can participate in polls)

a) OpenAI and news publishers can co-exist.

b) OpenAI and news publishers can’t co-exist.


For the Right:

How Modi government’s new criminal laws drastically increase police powers

For the Left:

In all the tumult of Bihar politics, one trend is clear