Reading time: three minutes

There’s something rather uncomfortable about the idea that a business could spend years building a website, investing in original content and developing a loyal readership, only to discover that the platform it uses to reach its audience may also want access to that content for its own commercial purposes.

That’s essentially the legal concern emerging from Meta’s proposed changes to its business terms.

The technology giant has found itself under scrutiny from the UK’s Competition and Markets Authority (CMA), following concerns that its proposed arrangements could give it extensive access to material published on business websites. And while Meta has since delayed the changes, the controversy raises a much bigger question about the relationship between AI companies and the businesses whose content helps power their technology.

This is no longer simply a discussion about AI innovation. It’s about commercial power, ownership and whether businesses have any meaningful say over how their work is used.

When having a business account could come at a cost

For many businesses, maintaining a presence on Facebook and Instagram is simply part of operating in the modern digital marketplace. These platforms provide opportunities to advertise, attract customers and maintain relationships with audiences.

But Meta’s proposed terms have raised concerns about what businesses might be agreeing to in return.

The issue centres on the possibility that Meta could access content from websites associated with business accounts, rather than being restricted to material that businesses have deliberately uploaded to its platforms.

Consider a publisher that’s spent considerable time and money producing original journalism. Its website may contain thousands of articles, photographs, investigations and other commercially valuable material. Under the proposed arrangements, publishers fear that simply maintaining a business presence with Meta could potentially expose this wider collection of content to automated scraping.

That’s a rather different proposition from allowing a social media platform to display a link to an article.

It also explains why publishers have been so concerned about the proposed changes.

Has Meta really removed the problem?

Meta initially attracted criticism because its proposed terms expressly referred to using collected website content to train its products.

Following the backlash, that specific reference was removed. However, the remaining wording still refers to using scraped content to develop, improve, research and maintain advertising and business products.

And this is where the legal uncertainty begins.

AI training doesn’t necessarily become irrelevant simply because a company removes the word “training” from its contractual wording. If the remaining permissions are sufficiently broad to encompass AI-related development, the underlying concern may remain.

The question is therefore not simply what terminology Meta uses, but what rights its terms actually seek to obtain.

This is an important distinction in contract law. Businesses should be able to understand the scope of the permissions they’re granting, particularly where those permissions could have significant commercial consequences.

Meta has maintained that it will respect website operators’ robots.txt instructions, which communicate preferences about automated crawling. However, publishers have questioned why such protections aren’t expressly reflected in the proposed contractual terms.

A public reassurance and a clearly enforceable contractual commitment aren’t necessarily the same thing.

Why the CMA has stepped in

The dispute has now moved beyond the publishing industry’s objections.

On 30 October 2026, the CMA announced that it had opened an investigation into Meta’s proposed terms under the Competition Act 1998.

The regulator will consider whether Meta holds a dominant position in a relevant UK market and, if so, whether imposing the proposed conditions could amount to an abuse of that position.

In its official announcement, the CMA explained that businesses could potentially face unfair trading conditions if accepting the proposed terms becomes a requirement for continuing to use Meta’s services.

This introduces an interesting competition law dimension.

A contractual agreement isn’t automatically fair simply because a business technically has the option to accept or reject it. Where one party has significant market power and the other relies on its services to reach customers, the commercial reality of that choice deserves closer examination.

That’s particularly relevant for smaller publishers, which may have limited alternatives when trying to reach large online audiences.

Sarah Cardell, chief executive of the CMA, has said the regulator is acting swiftly to investigate concerns raised by publishers and other businesses.

It’s important to stress that this remains an investigation; the CMA hasn’t found that Meta has breached competition law. Nevertheless, the regulatory scrutiny demonstrates how seriously the potential implications are being considered.

Publishers aren’t just fighting over copyright

It’d be easy to view this dispute purely as another argument about copyright infringement. However, that would overlook a much wider issue.

Publishers aren’t simply concerned about whether an individual article might be copied. They’re concerned about the commercial value of their entire content libraries and who gets to benefit from them.

Producing quality journalism requires investment in staff, research, technology and editorial expertise. If that material can subsequently be used to develop commercially valuable AI products without an appropriate licensing arrangement, publishers understandably want to know where their rights stand.

Sajeeda Merali, chief executive of the Professional Publishers Association, has argued that access to audiences shouldn’t come at the expense of publishers’ content.

That argument deserves attention. Businesses shouldn’t have to choose between maintaining their online visibility and retaining meaningful control over their intellectual property.

The legal position will, of course, depend on the particular material, the nature of its use and the relevant copyright exceptions. Public accessibility doesn’t automatically give AI companies unrestricted permission to use copyrighted material commercially.