Home › Guides › Bot ID Bills Compared: UK, US Congress and New York
✓ checked 24 October 2026 · sources: UK Parliament Bills API and the bill text as introduced, Congress.gov API and the GPO text of H.R. 9915, the New York State Assembly bill record, statements by Damian Hinds · method · caveats · download the data
Three bills now on the table, one in the UK, one in the US Congress and one in New York, would make bots say who runs them and what they want the content for. They share much of their wording, down to the list of “crawler, spider, fetcher, client, user agent, AI agent”. The differences are in who can enforce them and what happens to a bot that hides. The UK bill is the only one that lets a court order internet providers to block a repeat offender, and the only one with a public register of bots. New York’s version has passed both chambers.
bills compared: 3UK second reading: 27 Nov 2026US max penalty per violation: $53,000
Why this matters now
Damian Hinds’s Automated Online Software (Access and Transparency) Bill was first read in the Commons on 17 June 2026, and its text was published on 13 October as Bill 25. It was due a second reading on 16 October, but Hinds wrote on 20 October that it “was not reached” because of extended debate on an earlier bill. The Bills API now lists the second reading for 27 November 2026. On 14 October, UK national and regional news publishers, joined by a large website and a web infrastructure company, signed a joint letter asking the Prime Minister to back the bill. Private members’ bills rarely become law without government support. Hinds has also said he hopes the government might adopt the bill’s principles in a bill of its own.
The three bills side by side
| Provision | UK: Automated Online Software (Access and Transparency) Bill | US: Stealth Bot Prohibition Act (H.R. 9915) | New York: Stealth Crawler Prohibition Act (S9934-A / A11292) |
|---|---|---|---|
| Status on 24 October 2026 | Commons second reading listed for 27 November 2026; not reached on 16 October | Referred to the House Energy and Commerce Committee on 23 July 2026 | Passed the Senate on 2 June and the Assembly on 5 June 2026; not yet delivered to the governor |
| Sponsor | Damian Hinds (Conservative), private member’s bill | Rep. Laurel Lee (R-FL) and 4 cosponsors | Sen. Michael Gianaris (S9934-A); Assembly companion A11292 |
| Sites protected | Any website | Any website, digital platform or online service | Covered news sources only |
| What a bot must disclose | User-agent string, the bot’s tasks, operator identity and contact details, and the purpose and intended use of the data | Identity, including a valid and accurate user-agent string, and purpose and intended use of the content | User-agent stating the software, its version and the company behind it, plus all uses the content could be put to |
| Uses named | Text and data mining, search indexing, inferencing, AI training, fine-tuning and retrieval-augmented generation | Text and data mining, search indexing, inferencing, AI training, fine-tuning and retrieval-augmented generation | All uses and purposes |
| When it is unlawful | Any failure to disclose, or false or misleading information | Stealth access reasonably likely to damage, impair or burden a site, or disguising a bot as a human for generative AI | Stealth access that would damage, impair or burden a news source or cause it economic harm |
| Who enforces | Website operators in civil court; Ofcom compliance notices | Federal Trade Commission; state attorneys general | State attorney general |
| Penalties and remedies | Injunctions, damages, access bans, and orders for UK internet providers to block repeat offenders | Civil penalty of up to $53,000 per violation, inflation-adjusted | Civil penalties of up to $15,000 per day per violation |
| Public register | Yes, kept by Ofcom | No | No |
| Identifying unknown operators | Not specified | Not specified | Pre-action subpoenas to internet and network providers |
| Reach outside the jurisdiction | Explicitly yes | Not specified | Not specified |
| Comes into force | Two months after passing | 180 days after enactment | 90 days after becoming law |
What stands out
- The UK bill covers every bot visit, not only harmful ones. Under section 3, a bot must give its identity and purpose whenever it requests access to a website. The US and New York bills make hidden access unlawful only when it is likely to damage, impair or burden a site, or, in New York, cause a news source economic harm.
- Who is a bot operator. The UK text says a person is not an operator “by reason only of giving an instruction to a bot made available by another person (including by providing a website address, query or prompt)”. So the duty falls on companies running AI agents, not on the people typing prompts.
- Proxy networks are part of the argument. In the speech he had planned to give, Hinds described a blocked bot returning “as hundreds of visitors (known as residential IP proxies), routed through ordinary people’s routers, smart TVs and phones without their knowledge”. None of the three bills names proxies. They regulate whoever operates the bot.
- Enforcement routes differ. In the UK, site operators could sue directly, and Ofcom could issue compliance notices after giving 14 days for representations. In the US, enforcement lies with the FTC and state attorneys general, and actions must be brought within 6 years. In New York, the attorney general can seek up to $15,000 a day per violation. Publishers can also ask a court for subpoenas to make internet and network providers identify an unknown operator.
- New York covers news only. To be a covered news source, a site must, among other tests, publish at least monthly and have at least 1,000 monthly users in New York. A separate New York bill, the AI transparency for journalism act (A8595-B / S8331-A), which concerns training-data disclosure, has been in committee since January 2026.
- A drafting slip. The UK bill as introduced refers to “the public register kept under section 5”, but the register is created by section 4. Bills are often corrected in committee.
For the technical standards that would let bots prove who they are, see our Web Bot Auth tracker. For court cases and existing law, see the scraping law hub, and for how publishers are campaigning, publisher lobbying on AI crawlers.
Method
On 24 October 2026 we pulled the UK bill’s record, stages and publications from the UK Parliament Bills API. We read the as-introduced text (Bill 25, 2026–27) on publications.parliament.uk. For H.R. 9915 we used the Congress.gov API for status and cosponsors (5 members in total: 3 Republicans and 2 Democrats) and the GPO text as introduced. For New York we used the Assembly’s bill record for summaries, actions and the text of S9934-A. Each table row paraphrases the bill text, and quoted phrases are verbatim. Contains Parliamentary information licensed under the Open Parliament Licence v3.0.
Caveats
- All three are bills, not law. Text can change, and the UK bill is a private member’s bill with limited debating time.
- This page summarises the legislation and is not legal advice.
- “Not specified” means we found no express provision in the text. It does not mean the point could not arise under other law.
- The New York record shows no delivery to the governor as of 24 October. The bill would take effect only if signed.
Download the data
Cite as: ProxyPicker, “Bot ID Bills Compared: UK, US Congress and New York”, 24 October 2026, https://proxypicker.com/bot-identification-bills-uk-us-ny/.
Press & data requests: email [email protected] for interviews, the raw data behind this page, or corrections. See also our contact page.