Artificial intelligence is one of the fastest-growing, most exciting industries in America. Many in the tech industry are confident that as AI continues to improve, it will become increasingly important in our everyday lives.

But its growth has come at a cost — and if we’re not careful,  AI’s expansion could end up crippling other critical sectors of the American economy.

Big Tech’s dirty secret is that the success of its AI tools has been almost entirely built on theft.

These companies are scraping enormous amounts of copyrighted content, without permission or compensation, to fuel their AI products — and in the process dangerously undermining content creators’ businesses.

Instead of paying for access to copyrighted material — everything from magazine columns to President Trump’s own book “The Art of the Deal” — most AI companies have made the conscious choice to steal it instead.

They argue that all content, even content registered for protection with the US Copyright Office, should be considered “fair use” when used to build and operate AI models.

To gather the data that powers their large language models, Big Tech companies have consistently bypassed paywalls, ignored websites’ directives asking users not to copy material, and worse.

Meta, for instance, used illegal Russia-based pirate site LibGen to copy the contents of at least 7.5 million books to train its Llama AI model — an egregiously unlawful, copyright-violating workaround.

Relying on questionable sources for AI training purposes poses a variety of serious problems, perhaps even to US national security.

Recently an online data watchdog found that many of the most popular AI chatbots have absorbed millions of articles designed to spread Russian propaganda and outright falsehoods.

Now infected by a Russian disinformation network known as “Pravda,” these chatbots, including Grok, ChatGPT and Gemini, mimic Kremlin talking points when asked about certain topics — and spread false narratives about 33% of the time.

Content creators, meanwhile, face existential problems.

In addition to seeing their content stolen for training purposes, publishers are now forced to watch as Big Tech companies make billions using that stolen content in ways that directly compete with publishers’ business models.

With retrieval-augmented generation, in which an AI model references outside sources before responding to user inquiries, many AI products now give users real-time information, pulled directly from recently published news articles.

Those same AI companies run ads against that content — generating revenue that should belong to those who invested in its creation.

A user who gets all the information contained within an article directly through an AI chatbot has almost no reason to click through to the original text — much less to buy a subscription, if the item is behind a paywall.

The data on this is clear:  AI chatbots drive referral traffic at a 96% lower rate than traditional Google search, an already shrunken relic of the past.

For every 1,000 people using an AI chatbot, fewer than four will click through to see the original source of the information they’re reading.

As AI replaces traditional search for many users, this drop in referral traffic will cut deeply into both subscription and advertising revenue for publishers — depriving them of the funds they need to produce the content consumers (and AI companies) rely on.

AI companies are lobbying to legitimize this behavior, but Washington should take care.

Tilting the scales in Big Tech’s favor will undermine centuries of intellectual-property protections that have paid tremendous dividends for the United States, giving us countless advancements — and a competitive edge on the world stage.

Blessing the theft of American content would instantly erode our country’s foundation as an innovation powerhouse.

The news media industry supports a balanced approach. Many publications and journalists, in fact, now use AI to better serve their customers.

But it’s important to develop AI products responsibly and in coordination with the creators of the content they use, with the long-term sustainability of both AI companies and creators in mind.

If AI companies’ theft drives creators out of business, everyone ends up worse off.

To protect their work, over a dozen members of the News/Media Alliance recently sued Cohere, Inc., a growing AI company, for unauthorized use of their content.

They joined a number of other publishers, including News Corp and The New York Times, that are suing various AI companies to enforce their rights. 

Some in Big Tech are clearly beginning to recognize the problem with unfettered content theft.

We’ve seen a rapid proliferation of licensing agreements, in which AI companies pay publishers to use their content, over the last year. A News/Media Alliance collective is currently licensing content at scale. 

But without reinforced legal protections against content theft, bad actors will continue to exploit publishers and creators — undermining America’s creative industries to further tech’s own commercial interests. 

Danielle Coffey is president and CEO of the News/Media Alliance, which represents more than 2,200 publishers nationwide.