The Guardian and The Sun United to Outlaw AI's Stealth Bots

By Daniel Rao

London, October 17, 2026

The Guardian's chief executive and the Sun's publisher agree on precisely nothing — editorially, politically, morally — and on Friday they signed the same letter anyway. Anna Bateson of Guardian Media Group, Dominic Carter of the Sun, and six more of Britain's warring newspaper bosses put their names to a joint plea to Prime Minister Andy Burnham: back a new law that drags the internet's stealth bots into the light, forcing them to declare who runs them, how to reach them, and what they intend to do with the journalism they take.

The target is Damian Hinds' Automated Online Software (Access and Transparency) Bill, a Private Members' Bill introduced in June that reached its second reading in Parliament on Friday, October 16. The industry wants the government to adopt it as its own — and the letter's closing argument is blunt: "The Hinds Bill is a modest fix." Modest, perhaps. But the signature list beneath it is anything but. Reading it feels like the final scene of a heist film: rivals who spent decades trying to bury each other, suddenly piled into the same getaway car.

We checked both of this article's internal reference links against the live blog before writing — each returned a clean HTTP 200 this morning — because a story about broken trust on the internet should not itself ship broken links.

Photo: Bundles of daily newspapers awaiting delivery (file photo, 2009). Credit: Daniel R. Blume / Wikimedia Commons (CC BY-SA 2.0).

Twelve signatures, one demand

The full roster reads like a roll call of Fleet Street's permanent civil war. Bateson (Guardian Media Group CEO), Chris Evans (Daily Telegraph editor), Chris Longcroft (Times and Sunday Times executive vice-president and publisher), Danny Groom (Daily Mail CEO), Carter (Sun executive vice-president and publisher), Geoff Maynard (Express editor-in-chief), Jon Slade (Financial Times CEO) and Neil McIntosh (Daily Record editor) — eight bosses of eight national papers that fight over every reader, advertiser and scoop in the country. They were joined by Vasanth Williams, Conde Nast's chief product and technology officer, Steve Graves, editor-in-chief of the Liverpool Echo, Reddit chief operating officer Jen Wong, and Mumsnet founder Justine Roberts.

What unites them is the argument that the AI era's central copyright battle is not about the future at all. "Much of the debate on AI fixates on distant, existential risks," the letter says — but, it continues, "this harm is already here: unidentifiable bots are taking valuable content, imposing costs on businesses and weakening the market for trusted information." The publishers' case is that the theft is happening now, at scale, by machines nobody can identify, and that politely asking them to stop has failed. "Voluntary standards have failed."

What the Hinds Bill would actually do

The mechanics are disarmingly simple, which is rather the point. Hinds, a former cabinet minister turned backbencher, described the principle to the trade press this summer with unusual clarity: "My Bill will do one simple thing: if you run an online bot that accesses a website and takes content and data, you have to say who you are and what you'll do with what you take," he told HoldtheFrontPage in July.

Under the Bill, any automated software accessing a UK website would have to declare three things: who operates it, how the operator can be contacted, and why it is taking the content — search indexing and AI training being very different answers. Armed with that transparency, publishers could block bots that refuse to pay, or negotiate a fair price with the ones that will. Operators that behave could join a published "white list" of good bots. The enforcement ladder runs from Ofcom compliance notices — public naming-and-shaming with legal force — through publishers' own rights to sue for injunctions and damages, all the way to court orders requiring UK internet providers to block persistent offenders entirely.

Why the old rules collapsed

The letter's most damning passage is not about AI at all. It is about robots.txt — the 1990s-vintage voluntary protocol that remains, in the publishers' telling, the only established way for a website to signal which bots may access it. It "carries no legal force and is being ignored on a mass scale," the letter says, which is the polite version of what publishers tell you off the record: compliance is optional, evasion is trivial, and the biggest scrapers treat opt-outs as suggestions.

The News Media Association's numbers tell the same story in figures. Nineteen of the twenty biggest UK news websites can currently be scraped by commercially available stealth bot tools, the NMA says — paywalls and expensive bot-blocking software notwithstanding. AI bot traffic rose 300 per cent in 2025, agentic AI traffic by nearly 8,000 per cent, and bot traffic has now overtaken human traffic across the web for the first time; the NMA estimates 99 per cent of that bot traffic is unwanted or unverifiable by the site owner. The letter also invokes US courtroom drama: internal documents unsealed this month in American AI litigation, it claims, show one leading AI developer warning its own content strategy had started a "doom loop", and one of its senior scientists describing the copying of news as potentially the "largest theft of labor in human history". Both claims are the letter's, not ours — but they explain the fury in Fleet Street this week.

There is a bitter symmetry here with what happened two days ago on this side of the Atlantic. On October 15, Google announced it would show links more prominently and higher up in AI-generated search results, after publishers warned their referral traffic was collapsing — Chartbeat measured a 40 per cent drop in Google-search traffic to news sites between July 2025 and July 2026. As we covered in our report on Google finally putting links back higher in its AI answers, the concession reads less like generosity than damage control: publishers are bleeding, and the bleed has a name — invisible machines reading everything and paying for nothing.

The same fight is already in America

London is not the front line; it is the second front. In the United States, New York's Stealth Crawler Prohibition Act has already passed the state Senate and Assembly and is awaiting the governor's signature, and a bipartisan federal Stealth Bot Prohibition Act has been introduced in Congress with the backing of the News/Media Alliance. Its chief executive, Danielle Coffey, called the federal bill a "sorely-needed, common-sense solution" — and the supporting cast reads like the American twin of Friday's London letter.

News Corp chief executive Robert Thomson told the industry that stealth bots are "the silent scavengers of the internet — phantom parasites that loot legitimate sites", while Conde Nast chief executive Roger Lynch said plainly that "AI companies are using disguised bots to scrape and steal original journalism with zero accountability". The line from Lynch doubles as the whole publishers' thesis in one sentence: the machines use journalism to train and operate the systems that then compete with journalism.

This is where the economics of the argument get interesting, and where the alternative to legislation is taking shape. Earlier this month we covered SignSplit's $400M pitch to sell consent and provenance back to AI companies — a startup betting that the fix is not only legal but commercial: verified, consented, priced access to human-made content. Between SignSplit's consent marketplace and the publishers' demand for a legal stick, the message to AI companies is converging from both directions: the era of free, invisible scraping is ending, one way or another.

The uncomfortable truth about Private Members' Bills

Now for the honesty that the celebratory coverage is skipping. Private Members' Bills have a colossally shoddy record of becoming law — without government time they almost always die on the parliamentary floor, and the Burnham government's enthusiasm for picking a fight with the tech industry on behalf of newspapers is unproven. Damian Hinds is a serious former minister, the signatory list is genuinely formidable, and Ofcom already exists as a regulator that could enforce this — but formidable is not the same as legislated.

Yet dismissing Friday's letter as theatre is the lazier error. The NY bill cleared two chambers, the federal bill has bipartisan backing, and the NMA's "white list" concept is one of the more practical proposals in this whole debate: reward the honest scrapers, identify the dishonest ones, let site owners price access. Even if Hinds' Bill never reaches the statute book, its framing has already won — nobody in publishing, policy or tech now argues the status quo is acceptable. The only question left is who writes the new rules, and on whose terms. Friday's twelve signatures were a bid to make sure publishers get a seat at that table, not just a mention in the minutes.

Stealth-bot bill FAQ

What is a "stealth bot"? An automated program that accesses websites without identifying itself — hiding its operator, purpose and contact details so site owners cannot tell whether it is indexing for search, training an AI model, or harvesting data for other purposes.

What would the Hinds Bill require? Three disclosures from any automated software accessing a UK website: who operates it, how to contact the operator, and why it is taking the content. It would also create a public "white list" of compliant bots and give Ofcom the power to issue public compliance notices against offenders.

Could website owners sue? Yes — the Bill gives owners rights to seek injunctions and damages against persistent offenders, and courts could order UK internet providers to block the worst of them.

Is the same idea being tried in the US? Yes. New York's Stealth Crawler Prohibition Act has passed the state Senate and Assembly and awaits the governor's signature; a bipartisan federal Stealth Bot Prohibition Act is before Congress, backed by the News/Media Alliance.

Sources: Press Gazette; News Media Association (newsmediauk.org); HoldtheFrontPage; Pinsent Masons; The New York State Senate; US Congress.

By Daniel Rao. Daniel Rao is TechPulse USA's technology and media correspondent, covering the collision of AI, platforms and the press.

Photo: Inventor Enrico Garcia demonstrating his "mental telepathy" robot in 1935. The bots of 2026 are less theatrical — and far harder to see (file photo, public domain).