1. What the Online Safety Act Is Actually Asking For

Once you understand the shape of the law, filling in the paperwork becomes common sense. For a chat site, the Act boils down to two sets of duties:

1. Illegal content duties (apply to every user-to-user service, in force since March 2025): You must assess how your chat could be used for illegal activity and take proportionate steps to reduce it, then remove illegal content quickly when you become aware of it. For chat, the big risks are grooming and child sexual abuse material (CSAM), intimate image abuse, harassment and threats, hate, fraud and scams, plus drugs, weapons, terrorism and the other priority offences.
2. Children’s safety duties (apply only if children are likely to access your service, in force since July 2025): If under-18s are likely to use your chat, you must assess the risk of them encountering content that is harmful to children and take proportionate steps to protect them. The most serious category, which Ofcom calls “primary priority content”, is pornography, and content encouraging suicide, self-harm or eating disorders. The next tier, “priority content”, includes bullying, abusive or hateful content, serious violence and dangerous challenges.

What every small chat site needs as a baseline. Ofcom’s illegal content codes have a short set of measures that apply even to small, low-risk services:

• A written illegal content risk assessment, kept on file.
• A named person accountable for online safety compliance. This goes in your internal records and doesn’t have to be published on your site (see the note below).
• A content moderation function, meaning someone who reviews reported content and takes illegal content down swiftly.
• An easy way for users to report content and make complaints, and a process for acting on them.
• Clear terms of service explaining how you protect users.

Larger services, and services that assess themselves as medium or high risk for particular harms, get extra recommended measures on top. For example, a service at high risk of grooming is expected to apply safety defaults for child accounts so strangers can’t freely message them or find them.

A note on CSAM reporting: since 7 April 2026, UK-based user-to-user services must report detected child sexual exploitation and abuse content to the National Crime Agency (NCA) through its industry reporting portal. Non-UK providers must report UK-linked content.

3. How World of Chat Messenger Supports Your Safety Measures

Your risk assessment needs to say what technical and moderation controls you have. World of Chat Messenger (woc-messenger) gives you a set of built-in tools you can cite. They support your compliance; they don’t achieve it on their own. Ofcom still expects you (or your moderators) to review reports and act on them quickly.

5. Step-by-Step: How to Fill Out Each Section of the Template

Ofcom requires your assessment to be “suitable and sufficient”, meaning it reflects your site, your features and your evidence. Don’t copy another site’s completed assessment. You don’t send it to Ofcom; you complete it, date it, keep it on file, and produce it if Ofcom asks.

Header & Top Details

• Website Name: your site’s name and domain.
• Date Completed & Next Review Date: the date you complete it, and a review date no more than 12 months later. Ofcom recommends reviewing the illegal content risk assessment at least annually, and the Act requires you to redo the children’s access assessment at least every year. You also need to reassess before any significant change to your chat (for example, turning on photo sharing or opening chat to guests).
• Named Responsible Person: Ofcom’s codes recommend that every service names an individual accountable for online safety. This is an internal record, so use your real name and role here. It does not need to appear on your website. Publicly you can stay as a site username with a role-based contact email, which matters on chat sites where banned users sometimes try to track down and threaten the owner (see the note in section 1).

Section 1: About Our Website and Chat Service

Describe what your site actually is and what the chat can do. Be specific: a members-only chat on a classic car forum has a very different risk profile from an open stranger-chat site with photo sharing. The features you list here (private messaging, photo sharing, guest access, user profiles) are exactly the “risk factors” Ofcom’s Risk Profiles ask you to consider.

Section 2: Children’s Access Assessment

Work through the two Section 35 questions honestly and write down your evidence. This is a record of your reasoning, not a magic formula that gets you out of the children’s duties.

• Can children access the service? Unless you use age verification or age estimation that genuinely stops children getting in, write “Yes”. An 18+ rule or self-declaration doesn’t change this answer.
• Do a significant number of children actually use it? Give your evidence: your audience, what moderators and reports show, sign-up data, any users removed for appearing to be under 18.
• Is it likely to attract a significant number of children? Consider your topic, design and how people find the chat (for example, open guest chat reachable from search engines).
• Your age policy (if you have one): if you’ve chosen to run an 18+ community, say so and explain how you handle users who appear to be underage. Present it as a safeguard, not as proof that children can’t get in.
• Your conclusion: either “The child user condition is not met, so our service is not likely to be accessed by children. We will reassess within 12 months, or sooner if we see evidence of children using it,” or “Our service is likely to be accessed by children, so we have completed a children’s risk assessment and apply the protections set out below.”

Section 3: Protecting Children from Harmful Content (if children are likely to access the site)

If you concluded children are likely to access your chat, record how you protect them. For most chat sites this means:

• Your rules ban primary priority content (pornography; content encouraging suicide, self-harm or eating disorders). Keeping these banned, and removing them swiftly, is what keeps you outside the situations where Ofcom’s codes recommend age-check software.
• How harmful content is found and removed: user reports with 15-message context, your Chat Monitor, how often you check it, and how quickly you act.
• Explicit photos: the server-side scanner blurs likely-explicit photos as an extra layer. Or, if you’ve turned photo sharing off, say so.
• Bullying and abuse (priority content): blocking, reporting and bans.
• Grooming: how you stop adults using the chat to target children, such as monitoring reports of sexualised or age-related messages, removing offenders, and, if your risk is high, limiting who can privately message accounts you believe belong to children.

Section 4: The 17 Kinds of Priority Illegal Content

Ofcom asks you to assess each of the 17 kinds of priority illegal content separately, plus any other illegal content likely on your service (such as cyberflashing). For each, give a risk level of Negligible, Low, Medium or High, based on how likely it is on your site and how severe the impact would be, then note your evidence and mitigations.

Rate honestly, not optimistically. Don’t write “Negligible” in every box. Ofcom knows that harassment, abuse and scams are everyday problems in live chat, while terrorism or human trafficking are far rarer on a hobby site. Here’s how to approach the rows:

• Child sexual exploitation and abuse (grooming and CSAM): Ofcom’s Risk Profiles flag direct messaging, the ability to contact strangers, image sharing and anonymous or guest users as key risk factors for grooming, and a chat service will have several of these. Don’t rate it Negligible just because your rules say 18+. What to write: your relevant features; how grooming would be noticed (reports with context, moderators watching for sexualised or age-related talk); blocking; bans; photo scanning; and your process for removing CSAM, preserving evidence and reporting it to the NCA (or IWF and police).
• Intimate image abuse and extreme pornography: Relevant wherever users can send photos. Note whether photo sharing is on, the scanner and blur, reporting, blocking and bans. Also consider cyberflashing (sending unsolicited sexual images), a non-priority offence you should still cover.
• Harassment, stalking, threats and abuse, and hate: Often Low or Medium for an active chat. These are the most common real-world problems. Point to blocking, reporting, bans and how quickly you act.
• Fraud and financial offences: Scam, phishing and romance-fraud messages are common in open chats. Point to rate limits, duplicate-message filters, rules against solicitation, reporting and bans.
• Encouraging or assisting suicide: Note that your chat has no algorithmic feed, rules forbid it, moderators remove it, and you signpost users in distress to support (e.g. Samaritans on 116 123).
• Controlling or coercive behaviour and sexual exploitation of adults: Usually Low for a hobby site, but think about private messaging. Note blocking, reporting, and bans for solicitation or exploitation.
• The rest: terrorism; drugs and psychoactive substances; firearms, knives and other weapons; human trafficking; unlawful immigration; proceeds of crime; foreign interference; animal cruelty: For most small community sites these will be Negligible or Low. Briefly say why (e.g. “no marketplace or payment features”, “small community about [topic]”), that it’s banned by your rules, and that it would be removed, banned and referred to police if seen.

Section 5 & Section 6: What You Already Do & Honest Weak Points

In Section 5, confirm where your rules and terms are published and who moderates (you alone, or you plus volunteer moderators), and how often reports are checked. A moderation function that actually reviews reports and removes illegal content swiftly is one of Ofcom’s baseline expectations for every service.

In Section 6 (Weak Points), be candid. Ofcom’s guidance expects you to think about where your controls could fail. Good examples:

• Harmful text chat is mainly caught through reports and moderators, so it depends on someone checking regularly.
• Grooming can look like ordinary friendly conversation at first.
• You don’t use age verification, so you can’t be certain no under-18s are present.
• Automated photo scanning can miss images or flag innocent ones.
• A determined banned user could come back via a VPN.

Writing these down, along with what you do about them, shows that you’ve genuinely thought about how your chat works rather than ticking boxes.

Section 7: Responsibility, Serious Cases & Review

Fill in the responsible person and completion date. Keep the commitment that child sexual abuse content is removed immediately, the evidence preserved for at least a year, and the content reported to the NCA (see the NCA reporting section of this guide). If you have registered on the NCA portal, note the date. Save the finished document, and set a calendar reminder for your review.