whyframeshot - stock.adobe.com
The New York Times in multiyear licensing deal with Amazon
This is the first licensing agreement the publisher has made with a company in the GenAI era. It is indicative of a trend in media, as many publishers partner with AI vendors.
The New York Times is licensing its editorial content to a generative AI vendor for the first time.
The publisher revealed on Thursday that it has agreed to license its editorial content to Amazon.
The Times will bring editorial content through the multi-year agreement, such as news articles and materials from NYT Cooking and The Athletic, to Amazon products such as the Alexa voice assistant and the Amazon platform, which are infused with generative AI (GenAI) technology.
The agreement is a notable step for the Times, while it is currently suing AI vendor OpenAI and its backer, Microsoft.
In 2023, the Times sued OpenAI and Microsoft for copyright infringement, accusing the vendors of training their chatbots on the Times' articles without compensation.
The Times said its partnership with Amazon is consistent with its belief that "quality journalism is worth paying for."
While neither The Times nor Amazon has disclosed the amount the publisher is licensing its content for, the compensation likely is one of the key reasons the Times has agreed to this partnership, said Michael G Bennett, associate vice chancellor for data science and AI strategy at the University of Illinois Chicago.
"I bet that the dollar amount on this deal is a major driver in making Amazon, from the New York Times perspective, a good partner in this arrangement," Bennett said.
He added that Amazon's scale, which is much larger than OpenAI's, also makes it a good partner for The New York Times.
"The New York Times content is going to be making its way to users of Alexa; that's a huge market," Bennett continued. "There's also the fact that Amazon's infrastructure is probably growing quite rapidly."
A beneficial agreement
The partnership is also a way for the Times and other publishing companies to insist on compensation from vendors of AI systems that use their content. In recent years, publishers such as The Washington Post and Associated Press have partnered with GenAI model providers. In the GenAI arena, Amazon is known for its Amazon Bedrock large language model (LLM) platform and Titan foundation models.
These partnerships have come as many AI models scour the web for content.
"These trusted sources are becoming the currency of large language models," said Bradley Shimmin, an analyst with Futurum Group.
The model makers are also having a tough time because they’re running out of quality data to train their models, especially since some sources are locked behind paywalls. Moreover, if they get to the content past paywalls, they run into legal challenges.
“The model developers have already run into issues with these companies suing them, and so they’re looking to make deals,” said Chirag Shah, a professor of Information Science at the University of Washington in Seattle.
On the other hand, publishers face the challenge of either letting the model developers figure out a way to access their content or facing lawsuits, so partnerships are more sustainable, Shah said.
A survivor tactic
The internet is also changing, with search becoming more AI-centered. For example, Google last week revealed that it's introducing AI mode in Search and GenAI search vendors such as Perplexity and OpenAI make inroads in the search market Google has long dominated.
"We're walking in uncharted territories in terms of what the internet looks like," Shimmin said. "Is it going to be a dead internet where, as we see on Facebook, it might be dominated by AI-generated content, or is it going to be content created by experts in their fields?"
Therefore, for the Times, this partnership is more just than getting paid for its content; it is more about surviving, Shimmin added. “It’s a way of bringing their existing model forward without reinventing themselves," he said. "Instead, they're simply trying to take advantage of a broader shift in how people consume information."
Bennett said the Times is also considering its future, especially as it is still battling a fair use argument by OpenAI in the courts.
"It seems to be enabled by strategic thinking about how to increase the odds that The New York Times business model remains viable," he said.
It is also a way for the Times to pursue the fair use case against OpenAI in court by arguing that Amazon shows a right way to get access to their content, Shah said.
“This would also help with their lawsuit against OpenAI,” he said.
Some challenges
However, while the Times and other large publishers are able to procure such partnerships with tech companies, many smaller publishers can’t and may be hurt by LLMscrawling their sites, Shimmin said.
Moreover, consumer behavior could change, and consumers might stop visiting publishers' websites -- these types of partnerships might end up being detrimental, Shah said.
“There is actually a danger of a long-term demise of these publishers and journalistic avenues,” he said. He added that five or ten years from now, the model makers might have an upper hand. “All of these publishers are making short-term deals without thinking through the long term.”
Esther Shittu is an Informa TechTarget news writer and podcast host covering artificial intelligence software and systems.
