The value of journalism must be established in the AI era

Big Tech is building its latest technology on intellectual property and uncompensated use of expression, content, and data collected online and in databases.

Journalistic content, which is far more than just a collection of facts and is often gathered at great costs to the journalists who report the news, is indispensable to these new AI technologies.

The legal regulatory system has lagged behind recent rapid-fire developments in AI. By failing to enforce intellectual property rights, regulators have allowed a handful of companies to further entrench their dominance and develop technologies and business models that undermine the viability of entire sectors of the economy, including journalism.

The solution: News publishers, along with creative industries more broadly, must actively define the worth of their content and data by understanding how and why value is created throughout the generative AI process, from developing foundation models to powering real-time search, if they want to obtain fair compensation.

After decades of giving away their content for free and being held hostage to the power of social media and search platforms, news publishers are realizing that they need to be more proactive in the era of AI.

As AI companies rely on news content to train their large language models and make AI applications more relevant, publishers already contending with a precipitous decline in referral traffic and the continued monopolization of digital advertising by Big Tech are being exploited even further.

The journalism industry shed nearly 3,000 jobs in the U.S. alone and scores of publications closed over the past year, exposing the unviability of business models that supported news providers well into the 21st century. Publishers have seen referral traffic, already in decline since Facebook de-prioritized news, plummet even as they are trying to navigate the demise of cookies and the implications of AI for the future of their business.

Meanwhile, the tech companies propelling AI have enjoyed revenue growth and valuations that have turned them into the world’s most valuable companies with market capitalizations of more than a trillion dollars each.

This disconnect can be traced back to the damage tech corporations have wrought on news publishers by cannibalizing their original content and data, displaying them in their search results or social media feeds, and then diverting advertisers, readers, and potential subscribers away from the news sites themselves.

This reduces revenue from subscriptions, advertising, licensing, and affiliates, undermining not just the ability to produce quality journalism but also the industry’s underlying business model.

To adapt their business models for the AI era, news publishers need to demand their rights and work collectively to put a figure on the value of journalism to artificial intelligence systems and assess the threat posed to future revenue and business models.

Journalism content can serve as rich, diverse data that improve the accuracy and reliability of AI models while helping them better understand and interact with the world, particularly as synthetic media becomes more prominent online.

Journalism provides ongoing value because of its quality, timeliness, and empirical grounding, and it could become even more valuable as the amount of AI-generated content increases.

Access to human-created, high-quality content that is a relatively accurate and timely portrayal of reality, like journalism, is an important input for machine learning models.

News outlets must therefore consider how to optimize revenue streams and assert their pricing autonomy throughout the AI value chain. They will need to figure out how to unlock the value of journalism by adopting sophisticated and dynamic compensation frameworks and pricing strategies for news content in various parts of AI systems and applications.

Regardless of a handful of voluntary agreements, policymakers should explore statutory licensing and taxing generative AI firms to create a compensation fund that rights holders could apply for.

AI companies claim that it would be impossible to license data used in foundation models and compensate rights holders as if that should absolve them of the responsibility to do so.

But acquiescing to this stance means that we are prioritizing one business model over another. We are favoring a business model based on the pervasive theft of intellectual property by the wealthiest companies in the world over the business model of journalism.

How we decide to allocate intellectual property rights, and what we decide about how fair use does or does not apply to developing and training artificial intelligence systems, will have profound ramifications for business models in a variety of sectors and the further concentration of power in a handful of technology corporations.

Over the past nearly two decades, as tech companies like Apple, Amazon, Google, Meta, and Microsoft grew to become some of the most valuable companies in the world, the U.S. lost a third of its newspapers and two-thirds of its newspaper journalists. They cannot be replaced with AI.

Gone are the days of passive acceptance that enabled social media and search platforms to siphon off value from publishers and journalists without compensation.

We know that journalism is essential to democracy. Given AI’s well-established harms like the spread of misinformation during elections, we cannot say the same of generative AI.

Excerpted from a brief published by the Center for Journalism and Liberty. Courtney C. Radsch is director of the Open Market Institute’s Center for Journalism and Liberty and a fellow at the UCLA Institute for Technology, Law and Policy.