What counts as illegal content on a dating site?
Illegal content is content that amounts to a criminal offence covered by the Online Safety Act 2023. Section 59 of the Act defines it as content that amounts to a "relevant offence", and it covers words, images, speech and sounds.
There are two groups of relevant offence:
- Priority offences. These are listed in three schedules to the Act: terrorism (Schedule 5), child sexual exploitation and abuse, known as CSEA (Schedule 6), and a long list of other offences (Schedule 7).
- Other offences. Any other offence where the victim or intended victim is an individual. The Act excludes some areas, such as intellectual property and the safety or quality of goods.
On a dating service, content means more than photos. Profile text, private messages, voice notes and video are all in scope, because members can share them with other members.
This guide covers the UK rules in detail and the main US reporting duty. For the wider Act, see our guide to the Online Safety Act for dating sites. This page is general information, not legal advice.
Which kinds of illegal content matter most on a dating service?
Fraud, intimate image abuse, harassment and stalking, and grooming are the kinds Ofcom highlights for dating. Ofcom's page for dating and social discovery services says providers need to assess the risk of 17 kinds of illegal content, and it singles out those four.
Ofcom's summary of its illegal harms decisions lists the 17 kinds. They include terrorism, CSEA, hate, harassment and stalking, controlling or coercive behaviour, intimate image abuse, extreme pornography, sexual exploitation of adults, human trafficking, fraud and financial offences, and drugs and weapons offences. The list can change as Parliament adds offences, so check Ofcom's current guidance.
| Kind of illegal content | How it can show up on a dating service | First response |
|---|---|---|
| Fraud and financial offences | A fake profile builds trust, then asks for money, gift cards or an "investment" | Remove the account, warn members who were contacted, block the device and payment details |
| Intimate image abuse | A member shares or threatens to share a sexual image of someone without consent | Remove within the 48-hour limit, find copies, support the person pictured |
| Harassment and stalking | Repeated unwanted contact, or new profiles made after a block | Remove the accounts, stop re-registration, keep the evidence |
| Controlling or coercive behaviour | Threats and demands aimed at a partner or former partner | Remove, signpost support, keep the evidence |
| CSEA | An adult tries to contact a child, or shares an indecent image of a child | Remove, preserve, report (see below) |
| Sexual exploitation of adults and trafficking | Profiles that advertise paid sexual services or recruit people | Remove and report where you suspect exploitation |
Ofcom gives numbers for two of these. Its research found that 29% of adults who go online had experienced a romance or dating scam. It also cites a Thorn survey of young people threatened with exposure of sexual images, in which the perpetrator first made contact through a dating service in 11% of cases.
How do you decide whether a profile or message is illegal?
You ask whether you have reasonable grounds to infer that the content amounts to an offence. That is the test in section 192 of the Act, and it is lower than the standard a court uses.
The section says a provider must judge content on all relevant information that is reasonably available to it. What is reasonable depends on your size and capacity, and on whether people, software or both make the decision.
Reasonable grounds exist when two things are true:
- You have reasonable grounds to infer that every element of the offence is present, including the mental element, such as intent.
- You have no reasonable grounds to infer that a defence would succeed.
If both are true, you must treat the content as illegal. For content posted by a bot, the test looks at the person who may be assumed to control it.
In practice, most dating moderators never need to finish this test. If a profile breaks your own terms, for example by asking for money, you can remove it on that ground. Write terms that are at least as strict as the law, then train moderators to recognise the patterns in the table above. Our guide to a content moderation workflow covers the queue itself.
How quickly must illegal content come down?
Swiftly, and within 48 hours for reported intimate images. Section 10 of the Act sets three time duties.
- Priority illegal content. You must use proportionate systems to minimise the length of time it is present.
- Any illegal content. Once a person alerts you to it, or you become aware of it in another way, you must swiftly take it down.
- Intimate images. Subsection (3A), added by the Crime and Policing Act 2026 and in force since 29 June 2026, requires takedown as soon as reasonably practicable and no later than 48 hours after you receive an intimate image content report. It also covers content you identify as the same or substantially the same.
The 48-hour duty does not apply where you consider that the content is not an intimate image, or that the report did not come from the person pictured or someone acting for them. Our guide to image abuse on dating sites covers this in more depth.
The Act gives no number of hours for other illegal content. Set your own targets, and make sure the worst content, such as CSEA and threats, goes to the front of the queue.
What moderation and reporting measures does Ofcom expect?
Every in-scope service needs a way to review suspected illegal content, take it down swiftly, and receive complaints. Larger and riskier services need written policies, targets and trained staff as well.
Ofcom's summary of decisions shows which measures in its illegal content codes apply to which services.
| Measure | All services | Large, or smaller and multi-risk |
|---|---|---|
| A moderation function that reviews and assesses suspected illegal content | Yes | Yes |
| A moderation function that allows swift takedown | Yes | Yes |
| A complaints system that is easy to find and use | Yes | Yes |
| Action on complaints about suspected illegal content | Yes | Yes |
| Internal content policies, performance targets and prioritisation | No | Yes |
| Resourcing and training for moderators | No | Yes |
| A dedicated channel for trusted flaggers to report fraud | No | Large services at medium or high risk of fraud |
Automated hash matching for child sexual abuse material applies to large services at medium or high risk of image-based material, and to high-risk services above a threshold of "700,000 monthly active UK users" or that offer file storage and sharing.
Your terms of service must also say how members are protected from illegal content, describe any proactive technology you use, and be applied consistently.
What must a dating site report to the authorities?
Child sexual abuse material must be reported, and the route depends on where you are based. Most other illegal content has no general duty to report, though you may choose to.
UK. Section 66 requires UK providers to have systems that, as far as possible, report all detected and unreported CSEA content to the National Crime Agency. Providers outside the UK must report UK-linked CSEA content. Legislation.gov.uk shows the duty in force for user-to-user services since 7 April 2026, for content detected on or after that date.
US. Under 18 U.S.C. 2258A, a provider that obtains actual knowledge of apparent child sexual abuse material must report it to the CyberTipline of the National Center for Missing and Exploited Children as soon as reasonably possible. The statute does not require you to monitor members or search for material. A report counts as a request to preserve its contents for one year. Fines for knowingly and wilfully failing to report start at up to $600,000 for a first failure by a provider below the statute's 100 million monthly threshold.
For requests that come the other way, see our guide to law enforcement requests.
Who deals with illegal content on a white label dating site?
The platform that runs moderation and controls who can join carries most of the work, and the brand partner passes on anything it sees. Agree the split in writing before launch.
On Dating Partners, our sister company run by the same group as this site, the platform handles moderation, fraud screening and support across the shared network. It lists four layers of member checks: identity checks, AI scam detection, human review of every profile and ongoing behaviour monitoring. Partners handle brand, domain and marketing. Our safety and moderation page explains what the platform does.
A shared network does not suit every brand. If you want your own moderation rules or your own team reading reports, you need your own platform, and the duties above become yours.
Ask any provider:
- Who holds the illegal content risk assessment, and can I read a summary?
- What are your takedown targets, and how do you meet the 48-hour intimate image rule?
- Who reports CSEA content, and to which agency?
- If a member emails my brand about illegal content, where do I send it?
Questions
Is rude or offensive content illegal content?
Not by itself. Content is illegal only when it amounts to a criminal offence. You can still remove offensive content under your own terms of service, and most dating brands do.
Does a dating site have to scan private messages?
The Act asks for proportionate systems, not for every message to be read. Ofcom's automated detection measures for child sexual abuse material apply to larger or higher-risk services, and the US reporting law says it does not require a provider to monitor or search.
Are scam profiles illegal content?
They can be. Fraud and financial offences are one of the 17 kinds of priority illegal content in Ofcom's guidance, and Ofcom names romance fraud as a key risk on dating services.
Do we have to tell a member why their content was removed?
Ofcom's codes include measures on handling complaints and appeals about moderation decisions. A short notice that names the rule broken and explains how to appeal is good practice for a service of any size.
Who decides whether content is illegal: us or the police?
You do, for the purpose of your duties. The Act asks the provider to judge content on the information reasonably available to it, and it does not ask you to wait for a conviction or a police decision.
Sources
- Online Safety Act 2023, section 59: illegal content (legislation.gov.uk), checked 21 November 2026
- Online Safety Act 2023, section 10: safety duties about illegal content (legislation.gov.uk), checked 21 November 2026
- Online Safety Act 2023, section 192: providers' judgements about the status of content (legislation.gov.uk), checked 21 November 2026
- Online Safety Act 2023, section 66: requirement to report CSEA content to the NCA (legislation.gov.uk), checked 21 November 2026
- Ofcom: Dating and social discovery, know the online safety risks, rules and how to comply, checked 21 November 2026
- Ofcom: Illegal harms statement, summary of our decisions, checked 21 November 2026
- 18 U.S. Code 2258A: Reporting requirements of providers (Cornell Legal Information Institute), checked 21 November 2026
- Dating Partners partner terms, checked 21 November 2026
What changed
- : First published. Replaces our older guide on illegal content duties.