UK Regulators Take On Meta Over ‘Exploitative’ AI Terms for Publishers

Meta is facing a fresh regulatory storm on both sides of the North Sea. On Friday, Britain’s Competition and Markets Authority (CMA) opened a formal investigation into whether the tech giant’s proposed new terms and conditions — which would let it use news publishers’ content for its own products, including its AI features — amount to an abuse of its dominant market position. The probe lands just days after a Norwegian court temporarily barred Meta from enforcing the same terms, following a challenge from ten media outlets.

What the terms would do

At the heart of the dispute is a set of updated terms that Meta had planned to introduce on November 20, before putting the rollout on hold. The new language would allow Meta to access and use content from news publishers’ websites to develop and provide its own products — a category that explicitly includes the company’s growing suite of AI features. Publishers argue that consenting is not really optional: with Facebook and Instagram serving as the primary channel through which thousands of UK businesses reach their audiences, walking away from the platforms is not a realistic choice.

The CMA says its investigation will examine whether Meta’s terms “unfairly harm news publishers and other content businesses,” and specifically whether the company “is abusing a dominant position by imposing exploitative terms on publishers and advertisers.”

A regulator moving fast

CMA chief executive Sarah Cardell signaled that the watchdog is treating the matter with urgency. “We’ve heard concerns from news publishers and other businesses and are acting swiftly to investigate whether Meta’s proposed changes may breach competition law,” she said in a statement.

The speed is notable. The CMA has spent recent years building out its digital markets toolkit, and this case gives it an early, high-profile test of how far UK competition law can stretch into the contentious territory of AI training data. If the regulator finds that terms of this kind constitute an abuse of dominance, it could reshape how large platforms negotiate with the content creators whose work feeds their recommendation engines — and, increasingly, their models.

Norway got there first

The UK is not acting alone. A Norwegian court this week temporarily banned Meta from enforcing the new user terms after ten media outlets argued the changes would threaten their rights to their own content. The ruling is an interim measure, but it signals that European courts are willing to step in quickly when publishers’ intellectual property is at stake.

For Meta, the coordinated pushback is an unwelcome echo of the broader copyright reckoning facing the AI industry. Publishers and authors worldwide have spent the past two years challenging the mass use of their work to train and power AI systems, arguing that platforms and model developers are extracting enormous value from content they did not create and do not pay for. Meta’s proposed terms would have made that extraction explicit — codifying, in fine print, a rights grab that publishers say leaves them with no meaningful choice.

Meta’s response

The company has struck a conciliatory note. A Meta spokesperson said the company has “already committed not to change our terms without consulting publishers and other stakeholders,” adding: “We welcome the opportunity to explain our approach to the CMA.”

That commitment to consultation may soften the immediate controversy, but it does not resolve the underlying question: whether a platform of Meta’s scale can unilaterally rewrite the rules of its relationship with the publishers who depend on it. The CMA’s investigation will now attempt to answer exactly that.

Why it matters beyond the UK

The stakes extend well past one company or one country. Every major AI developer is hungry for high-quality, human-authored content, and news archives are among the most valuable training material available. If regulators conclude that dominant platforms cannot fold content-use rights into take-it-or-leave-it terms of service, the decision could become a template for scrutiny in other jurisdictions — and a bargaining chip for publishers negotiating licensing deals elsewhere.

For the tech industry, the case is also a reminder that AI’s regulatory risks are no longer just about safety, bias, or misinformation. The competitive economics of who gets to use whose content — and on what terms — is fast becoming the frontline of tech policy. Britain’s competition watchdog has just made that abundantly clear.