Thirty-one UK websites today launched legally binding contracts that set out payment terms for AI companies scraping their content without permission - and give publishers a clear path to enforcement through the county court system, bypassing costly intellectual property litigation entirely.

The initiative, organised by the Movement for an Open Web (MOW), introduces what the group calls "Search-Only Contracts" (SOC). These are added to website terms and conditions as well as robots.txt files, and they prohibit large language models such as OpenAI's ChatGPT and Google Gemini from copying and repurposing content. The contracts set a typical fee of £500 per article for unauthorised scraping. According to MOW, any AI company that accesses a participating website and then uses its content in a generative AI service has entered into the contract and agreed to those payment terms.

The announcement, published June 15, 2026, opens a new tactical front in a conflict that has been building across the publishing industry for well over a year.

How the county court mechanism works

The technical and legal architecture of the Search-Only Contract is designed to be simple enough for small publishers to use without specialist legal counsel. According to MOW, establishing that a particular LLM has taken a publisher's content is achievable by questioning the chatbot directly and recording what it produces. That output becomes the evidence base for a formal invoice issued to the LLM operator.

If the invoice goes unpaid, the publisher can file a claim at Moneyclaim.gov.uk - the government's online small claims portal - at a cost of around £50. The claim is then heard at a local county or small claims court. For claims under £10,000, claimants routinely represent themselves. The entire process sidesteps the copyright and intellectual property frameworks that have complicated much of the existing litigation between publishers and AI companies.

Should a tech company ignore a county court judgment, enforcement escalates to debt recovery. A judgment creditor can instruct bailiffs to attend the debtor's UK premises and seize assets for auction. According to Press Gazette, which first reported the initiative, Google's Platform 37 headquarters at King's Cross in London - a facility valued at roughly $1 billion and due to open this summer - could be subject to such enforcement action. The article specifically noted the £5,500 Metronap sleep pod chairs installed at the site as an example of seizable property.

The simplicity is the point. As MOW co-founder and lawyer Tim Cowan said: "This is a straightforward approach to dealing with the complexities of payment for content online. By avoiding copyright and IP law a tried and tested mechanism is available to publishers wishing to assert their content ownership over AI harvesting."

Cowan continued: "Using the SOC, website owners can simply invoice AI businesses for the content that they have taken and - if they don't pay - pursue them through the county courts for quick and cost effective resolution."

The rate structure itself matters. At £500 per article, a publisher whose content has been scraped thousands of times faces a situation where multiple claims, each small enough for county court, aggregate into significant potential liability for AI operators. Cowan framed this as a market-creation mechanism: "The SOC for the first time puts a price on content usage by AI and - in doing that - creates a viable marketplace for publishers' content assets. The AI companies can't ignore this: with a critical mass of support from publishers they are going to be forced to the negotiating table or else will face an ever-growing volume of court summons."

The scraping problem in numbers

The scale of the problem for publishers is not abstract. Candr Media chief executive Chris Dicker, whose company owns Trusted Reviews, disclosed that AI bots currently scrape the site approximately 200,000 times per day. That figure includes all major LLMs - OpenAI, Google, Meta, and others - and continues despite multiple attempts to stop them. Trusted Reviews is one of the 31 sites to have already added the Search-Only Contract to its terms and conditions.

Trusted Reviews previously crashed on August 16, 2025, when AI bots scraped the site 1.6 million times in a single day. According to figures Dicker shared publicly, that surge produced just 603 actual human visitors from generative AI platforms - a conversion rate of 0.037 percent. The bandwidth cost of those scrapes is borne by the publisher; the traffic value flows to the AI operator.

Dicker said: "For too long the AI companies have had a free hand to steal our content, traffic and IP for their own gain. With the Search-Only Contract we finally have a legal means to assert our ownership and take back control. The AI companies agree to the terms the moment they access our site, and if they breach them by using our content in their services, we issue a demand for payment. No negotiation, no grey areas. You took it, you pay for it."

The broader picture is consistent with the publisher data. Unauthorised scraping increased 40 percent from Q3 to Q4 2024, with robots.txt compliance declining significantly. Over 35 percent of the world's top 1,000 websites now block OpenAI's GPTBot web crawler, a seven-fold increase from August 2023. Yet blocking alone has not stopped the scraping for sites like Trusted Reviews.

Why robots.txt has not been enough

The Search-Only Contract explicitly addresses the inadequacy of existing technical controls. The robots.txt protocol - a text file placed at a website's root directory to signal crawler permissions - is widely deployed but carries no legal weight on its own. AI companies have treated it as advisory.

The IAB Tech Lab, in a framework published in June 2025, documented that robots.txt compliance among AI crawlers had declined significantly, even as the volume of scraping rose. The Ziff Davis lawsuit against OpenAI - filed in April 2025 in Delaware federal court - alleged that OpenAI continued scraping Ziff Davis properties including CNET, IGN, and Mashable despite explicit robots.txt disallow directives.

According to MOW, the SOC "seeks to beef up existing robots.txt notices on websites, which are currently widely ignored by generative AI companies." The contract converts a technical request into a legal obligation. When an AI bot accesses a website, it encounters the terms of the contract - and according to MOW's interpretation, accepting those terms is the act of access itself. This framing would allow a publisher to argue breach of contract rather than copyright infringement, which is a substantially simpler legal claim.

The SOC is compatible with existing website terms and conditions. It is hosted online at m4ow.uk/socw/2.txt, with implementation guidance at m4ow.uk/socw/. Publishers link to it from their robots.txt file, their website footer, or both. MOW says the contract is "machine readable" and does not conflict with existing content licensing agreements.

Specialist sports publisher F-At has deployed the contracts across three properties: road.cc, off.road.cc, and ebiketips.co.uk. F-At head of content Tony Farrelly said: "None of Claude, Gemini, ChatGPT or AI Mode can ride a bike nor tell you from first-hand experience how it handles in traffic or on a mountain road, but that hasn't stopped them taking the words of those that can, without permission or payment, putting their livelihoods at risk in the process."

Farrelly added: "It's clear that if independent and specialist publishers want to get a fair deal from Google and the AI companies then we're going to have to band together and do it ourselves as neither Government nor regulators have shown the will to hold the tech companies meaningfully to account while they steal our content and reduce, our revenues and most importantly the jobs of the thousands of talented people that work across the independent media sector."

Trade bodies back the mechanism

The Search-Only Contract has drawn formal support from two of the UK publishing industry's main trade associations: the Association of Online Publishers (AOP) and the Professional Publishers Association (PPA).

AOP managing director Richard Reeves framed the SOC within a broader argument about value exchange: "AOP continues to encourage all publishers to make use of the initiatives at their disposal to support the drive toward a fair value exchange. This is the only way to ensure the marketplace evolves in a direction that sees publishers of original journalism appropriately compensated for the use of their hard-earned, cost inducing intellectual property."

Reeves described the contract's technical design as a practical advantage: "The Search-Only Contract is one of these initiatives. It is simple to deploy, machine readable, and sitting within the robots.txt protocol, employing a frictionless common legal language that creates consistency, while complimenting existing T&Cs publishers may choose to utilise."

PPA chief executive Sajeeda Merali characterised the contract as a tool for bargaining power rather than a singular solution: "The Search-Only Contract is an interesting approach that seeks to give publishers additional choices over the terms on which their content can be used. While there will be different approaches to licensing and monetisation, what matters is that PPA members have options and bargaining power in the AI ecosystem."

The involvement of trade bodies is significant in terms of coordination. A fragmented publisher response - with individual outlets pursuing individual claims through individual courts - is harder for AI companies to manage than an industry-level framework. The county court mechanism scales precisely because it does not require centralisation.

The CMA connection

The Search-Only Contract's launch coincides with, and explicitly relies on, a regulatory development at the UK Competition and Markets Authority. In early June 2026, the CMA imposed its first binding conduct requirement on Google under the Digital Markets, Competition and Consumers Act 2024, requiring the company to give publishers controls over whether their content appears in AI Overviews and AI Mode.

Google responded by launching a new toggle in Search Console allowing website owners to opt out of AI features - a change that had been described in February 2026 by a senior Google executive as "a huge engineering project." The CMA's requirement includes an explicit anti-retaliation clause: Google cannot downrank publishers who choose to use AI controls.

According to MOW, the SOC is directly compatible with this regulatory framework. The group said: "The CMA's conduct requirements also set out that Google cannot retaliate against publishers for using AI controls by harming their prominence in search, ensuring that implementation of the SOC is risk free for publishers."

Cowan cited the CMA ruling as strengthening the SOC's position: "The CMA's recent ruling on Google's AI harvesting further strengthens the position of the Search-Only Contract. It demonstrates that Google's theft of publisher content without permission is unacceptable and it forbids Google from retaliating against those who choose not to give their content away for free. It places a requirement on Google not to retaliate."

Publishers can, according to MOW, use the SOC in two ways under the CMA framework: opt out of Google's AI services entirely while controlling other AI access through the contract, or remain opted in to Google's AI features but set the commercial terms of that access through the SOC's clauses.

Why this matters for the marketing community

The publisher-AI conflict is not peripheral to digital advertising. It sits at the centre of how the web's information economy functions - and how much publisher inventory exists to buy.

PPC Land analysis of Google's AI Overviews found that paid click-through rates on queries featuring AI Overviews fell 68 percent between June 2024 and September 2025. Google Web Search traffic to news publishers declined from 51 percent to 27 percent between 2023 and 2025. Chartbeat data shows small publishers lost 60 percent of search traffic in two years. Index Exchange data showed that 69 percent of publishers on its exchange experienced year-over-year declines in ad opportunities throughout 2025, with the average decline across that group reaching 14 percent.

Fewer publisher page views means less advertising inventory. Less inventory from independent publishers means more concentration of digital spend in fewer, larger properties. That dynamic has direct implications for advertisers seeking reach outside walled gardens.

The Search-Only Contract does not directly restore traffic. What it introduces is a commercial claim on the value that AI companies extract. Whether AI operators will pay, negotiate licensing deals, or contest the claims in court remains to be seen. But the mechanism changes the default: instead of content being freely available to LLMs at zero cost, a price is now attached. Cowan's argument is that a sufficient volume of claims forces negotiation. The same argument underpins the broader pattern of publisher litigation - from the New York Times lawsuit against OpenAI, to Ziff Davis's April 2025 case, to Anthropic's $1.5 billion copyright settlement in September 2025.

What is different about the county court route is speed and cost. Large intellectual property cases take years to resolve. A claim under £10,000 at a small claims court can move in weeks. If 500 publishers each file a £10,000 claim per month, the combined volume of paperwork - and potential judgment debt - is a different kind of problem than a single landmark case handled by specialist IP counsel.

One observer commenting on LinkedIn noted the strategic logic directly: "Tech companies are built to delay big legal battles, they're not built to handle 500 publishers each filing a £10k claim every month."

What comes next

The SOC is free to use. MOW says any website can add it by linking to the hosted contract from its robots.txt file and website footer. The group provides implementation guidance at the dedicated website. The 31 founding sites cover technology, cycling, and general interest content; the coalition is expected to grow as the mechanism is tested.

The next test is whether AI companies respond. Three options are available to them: pay the invoices, challenge the contractual mechanism in court, or enter licensing negotiations. A fourth option - ignoring the invoices and hoping enforcement does not follow - risks precisely the kind of county court judgments that the MOW framework is designed to produce.

For publishers outside the UK, the mechanism is specific to English law and English courts. Whether equivalent frameworks could operate in other jurisdictions depends on local contract law and the availability of accessible small claims procedures. The EU context is different, given ongoing regulatory proceedings and publisher antitrust complaints filed with the European Commission in June 2025 by the Independent Publishers Alliance and MOW.

The marketing and advertising community should watch how the county court strategy interacts with the CMA's conduct requirement over the coming months. If publishers begin filing claims in volume and AI companies face enforcement action at their UK offices, the pressure to formalise licensing markets will intensify - with consequences for how AI-generated content surfaces in paid and organic search, and how the inventory that supports programmatic advertising is sustained.

Timeline

Summary

Who: The Movement for an Open Web, co-founded by Tim Cowan and James Rosewell, backed by the Association of Online Publishers and the Professional Publishers Association. Founding participants include Candr Media (Trusted Reviews) and specialist publisher F-At (road.cc, off.road.cc, ebiketips.co.uk). The contracts are directed at AI companies including OpenAI (ChatGPT) and Google (Gemini).

What: The launch of Search-Only Contracts - legally binding terms added to website robots.txt files and terms of service - that prohibit AI companies from copying and repurposing publisher content, set a standard fee of £500 per article for unauthorised scraping, and establish a county court enforcement mechanism for collecting unpaid invoices.

When: Announced on June 15, 2026. The 31 founding sites have already added the contracts to their terms and conditions.

Where: The mechanism operates under English law and is enforceable through local county courts in the United Kingdom. The SOC is hosted at m4ow.uk/socw/2.txt, with implementation guidance at m4ow.uk/socw/. Enforcement against non-paying AI companies could extend to their UK offices, including Google's King's Cross headquarters.

Why: Existing technical controls - primarily robots.txt files - have been widely ignored by AI companies. Copyright and intellectual property litigation is slow and expensive. The county court small claims route offers a faster, lower-cost enforcement path that individual publishers can pursue without specialist legal representation, at a filing cost of approximately £50 per claim. The initiative also responds directly to the CMA's June 2026 binding conduct requirement, which prohibits Google from retaliating against publishers who use content controls.