News industry divides over AI

Driving the news: Eight prominent regional U.S. newspapers joined the New York Times and other news organizations in suing ChatGPT parent OpenAI and Microsoft for copyright infringement earlier this week.

• The new suits add heft to the Times' claims. Until now, the Times was the only major newspaper to take legal action against AI firms for copyright infringement. Smaller or newer outlets, like The Intercept, Raw Story and AlterNet, have also sued OpenAI and Microsoft.

Several other large news publishers, including the Financial Times, the Associated Press and Axel Springer, have instead opted to strike paid deals with AI companies for millions of dollars annually, undermining the Times' argument that it should receive damage payments of billions of dollars.

• The regional newspaper lawsuit was filed in the same district as the Times' complaint. If the same judge is chosen to oversee both cases, the litigants could choose to combine their two complaints.

Reality check: The Times has some leverage in its lawsuit because the firms it sued were actively trying to strike a deal with the paper for months.

• The eight newspapers that sued OpenAI and Microsoft this week, and many other news companies, have not been in active discussions with those AI firms about licensing deals at all.
• AI firms may not need to license news content from every publisher, depending on the type and volume of content they need to train or inform their large language models.

The big picture: AI developers like ChatGPT maker OpenAI don't generally share details of the content they used in training their models.

• In most cases, they've simply said they are making fair use of "publicly available" data.
• But many publishers believe they're sending automated tools to scrape copyrighted material that they should be paying for.

Between the lines: News organizations have evolved along different paths in the digital era, leaving many newsrooms with conflicting objectives in negotiating with AI companies.

• Advertising-based news businesses depend on visitors sent via search, and they fear tech companies will use their content to refine AI-based services that will simply give users information without sending traffic.
• Licensing businesses, like the AP, have less to lose in striking a deal with AI firms, as most of their revenue is already derived from deals providing access to their content.

• Older news companies with many decades of content have archives that could be very valuable to AI firms looking to train their large language models, while newer firms that break lots of news could help AI providers deliver real-time data and insights.
• News publishers with large video archives, such as broadcasters and cable networks, have slightly more leverage than text-based businesses, because video archives are typically less available for bots to scrape.

What to watch: Right now, figuring out the most lucrative way to partner with AI companies is hard for publishers because there's no marketplace to help buyers and sellers agree on rates.

• Some publishers are looking to partner with companies like TollBit, which is building a marketplace to connect AI bots and scrapers with publishers' verified content for a dynamic fee.
• Fox Corp. is building a similar marketplace tool using blockchain technology to help publishers track when their content is being scraped so that they can better negotiate with AI companies.
• To hedge, the AP — as part of its licensing agreement with OpenAI — has negotiated the right to reset its payout terms if another publisher strikes a better deal with OpenAI in the future.

The bottom line: While the news industry divides its efforts, tech firms haven't waited for payment terms to be negotiated, either in courts or by the market — they've gone ahead and taken the data they need.