facebook-pixel

UK Online Safety Act: What It Means for Your Privacy in 2026

L
Lunyb Security Team
··10 min read

The UK Online Safety Act is one of the most ambitious pieces of internet regulation in the world, and in 2026 its impact on everyday users is finally becoming tangible. Age verification walls are appearing on major platforms, Ofcom is issuing its first enforcement notices, and privacy campaigners are warning that some provisions could reshape how anonymously we can use the web. If you live in the UK — or run a service that reaches British users — understanding this law is no longer optional.

This guide breaks down what the Online Safety Act actually requires, how it affects your personal privacy, what platforms must do differently, and the practical steps you can take to stay safe and private without falling foul of the new regime.

What Is the UK Online Safety Act?

The UK Online Safety Act 2023 is a wide-ranging law that places a statutory "duty of care" on online platforms to protect users — especially children — from illegal and harmful content. Ofcom, the UK communications regulator, enforces the Act and can fine non-compliant services up to £18 million or 10% of global annual turnover, whichever is greater.

The Act was passed in October 2023, but its provisions are being phased in across 2024, 2025 and 2026. By early 2026, the three main pillars — illegal content duties, child safety duties, and transparency obligations — are all live.

Who the Act Applies To

The law covers any "user-to-user" service or search service with a link to the UK, regardless of where the company is based. That includes:

  • Social media platforms (Facebook, X, TikTok, Instagram)
  • Messaging apps (WhatsApp, Signal, Telegram)
  • Video-sharing sites (YouTube, Twitch)
  • Forums, dating apps and gaming platforms
  • Pornography websites
  • Search engines
  • Cloud storage with sharing features

Small services are not exempt simply because they are small. Ofcom uses a tiered system based on risk and reach, but even niche forums can fall under some duties.

How the Online Safety Act Affects Your Privacy

The Act's privacy implications are significant and often underappreciated. While it is framed as a safety law, several mechanisms require platforms to collect, analyse or verify more information about you than before.

1. Age Verification and Age Estimation

Any service likely to be accessed by children — or that hosts adult content — must now use "highly effective age assurance." Ofcom's guidance makes clear that a simple tick-box ("I am over 18") is no longer acceptable. Accepted methods include:

  1. Photo-ID matching (uploading a passport or driving licence)
  2. Credit card checks
  3. Mobile network operator age checks
  4. Facial age estimation using AI
  5. Digital identity wallets

Each of these requires sharing sensitive data — biometrics, government ID, or financial details — either with the platform or a third-party verification provider. Even where data is "double-blind" (the verifier doesn't know which site you're visiting), you're creating a new data trail that didn't exist before.

2. Pressure on End-to-End Encryption

Section 121 of the Act gives Ofcom power to require services to use "accredited technology" to scan for child sexual abuse material (CSAM) in private communications. In theory this applies to encrypted messaging. In practice, the government conceded in 2023 that the technology to do this without breaking encryption doesn't yet exist, and the power will not be used "until it is technically feasible."

That stay of execution is reassuring but not permanent. Signal and WhatsApP have both said they would withdraw from the UK rather than compromise encryption. For now, your messages remain private — but the legal mechanism to change that sits on the statute book.

3. More Data Collection by Platforms

To meet their duties, platforms must proactively identify illegal content, assess risk, and keep records. That usually means more logging, more behavioural analysis, and more automated scanning of what you post. Even content that is perfectly legal may be flagged, reviewed by a human moderator, or retained longer for audit purposes.

4. Reduced Anonymity Options

Category 1 services (the largest platforms) must offer users the option to verify their identity and to filter out interactions with unverified accounts. This is voluntary from your side, but it creates a two-tier internet: verified users get full reach, unverified users risk being filtered out. Over time this could erode the ability to post anonymously about sensitive topics such as whistleblowing, mental health or abuse.

Key Duties on Platforms Under the Act

Understanding what platforms must do helps you anticipate how your experience will change. The table below summarises the main duties by category of service.

DutyApplies ToPrivacy Impact
Illegal content risk assessmentAll user-to-user and search servicesMore scanning and logging
Children's access assessmentServices likely accessed by childrenAge checks introduced
Highly effective age assuranceAdult content sites, high-risk servicesID or biometric data required
User identity verification optionCategory 1 servicesOptional ID link to account
Transparency reportingCategory 1, 2A, 2BAggregate data published
Content scanning (Section 121)Messaging services (dormant)Potential future encryption risk

The Age Verification Debate

Age assurance is where the Online Safety Act most directly touches ordinary users. From mid-2025 onwards, a growing list of UK-facing sites — including Reddit, X, Bluesky, Pornhub and even some Wikipedia-style communities — began rolling out checks.

What You're Actually Asked to Share

Depending on the method, you may be asked to:

  • Upload a photo of your passport or driving licence
  • Record a short video selfie for liveness checks
  • Allow facial age estimation from a still image
  • Enter credit card details for a nominal authorisation
  • Authenticate via your mobile phone contract

Risks to Be Aware Of

Even when a provider claims not to store your data, breaches happen. In 2024 and 2025 several age-verification vendors were targeted by criminals specifically because of the sensitive data they handle. If you must verify, prefer providers that use on-device processing, publish third-party security audits, and explicitly commit to deleting raw data within minutes of the check.

Practical Steps to Protect Your Privacy

You cannot opt out of UK law, but you can minimise the data you expose. Here is a practical checklist tailored to the post-Online Safety Act landscape.

1. Choose Privacy-Respecting Age Checks

Where a site offers multiple age-assurance methods, pick the one that shares the least. Facial age estimation (which returns only an estimated age band) is generally less revealing than uploading a full identity document. Digital identity wallets that disclose only "over 18: yes/no" are the gold standard.

2. Use Separate Emails and Accounts

Compartmentalise. Use one email address for verified, identity-linked accounts (banking, government, work) and separate aliases for forums, social media and newsletters. Services like Apple Hide My Email, Firefox Relay, or SimpleLogin make this almost effortless.

3. Harden Your Browser and DNS

Switch to a privacy-oriented browser such as Firefox or Brave, enable encrypted DNS (DNS over HTTPS) through a provider like Cloudflare 1.1.1.1 or Quad9, and install uBlock Origin. These steps reduce the metadata leaking from your device without requiring any special network tools.

4. Shorten and Audit the Links You Share

When sharing links publicly — on social media, in forum posts, or in support groups — long URLs can leak referral data, campaign tags and even account identifiers. A privacy-first URL shortener such as Lunyb lets you create clean, trackable-only-by-you links without handing analytics to a third-party advertiser. If you want to know more about how it compares to competitors, our honest review of Lunyb and our Rebrandly comparison walk through the differences.

5. Review App Permissions Regularly

Under the Act, platforms are incentivised to collect more signal about who you are. Periodically revoke microphone, camera, contacts and location permissions for apps that don't strictly need them. On iOS and Android, check which apps have accessed sensitive data in the last week.

6. Use Encrypted Messaging — While You Still Can

Signal remains the gold standard for private messaging. WhatsApp also offers end-to-end encryption. Both have stated they will leave the UK rather than weaken encryption, so continuing to use them signals demand for privacy-preserving services.

What the Online Safety Act Does Not Do

There are widespread myths about the Act. To put your mind at ease, here is what it does not require:

  • It does not require you to show ID to use the internet in general.
  • It does not ban anonymous accounts on social media.
  • It does not currently break end-to-end encryption.
  • It does not criminalise ordinary users for sharing lawful opinions, however unpopular.
  • It does not require platforms to hand over your personal data to the government on demand without a warrant.

The duties fall on platforms, not users. You can still be anonymous, still use encrypted apps, and still speak freely within the limits of existing UK law (which was already strict on defamation, harassment and incitement).

How the Act Compares Internationally

The UK is not alone, but its approach is distinctive. The table below puts the Online Safety Act in context.

JurisdictionLawAge ChecksEncryption Risk
UKOnline Safety Act 2023Mandatory for high-riskDormant Section 121 power
EUDigital Services ActRisk-based, not mandatedSeparate CSAM proposal pending
AustraliaOnline Safety Act 2021Trial underwayeSafety notices possible
USAState-level (e.g. Texas, Utah)Mandatory for adult sitesNo federal mandate

Looking Ahead: 2026 and Beyond

The next 12 months will be defining. Ofcom is expected to issue its first major fines, the first legal challenges will reach the courts, and the government's long-promised review of the Act's impact on small services and encryption is due. Expect further codes of practice covering AI-generated content, deepfakes and livestreaming.

For users, the trajectory is clear: identity will increasingly be tied to online activity, and privacy will depend more than ever on the choices you make about which tools you use. Being deliberate about your browser, your messaging app, your email aliases and even the link shorteners you trust is no longer a niche hobby — it's basic digital hygiene.

Frequently Asked Questions

Does the UK Online Safety Act require me to upload my passport to use social media?

No. Mainstream social media does not currently require identity verification for ordinary use. Age assurance is required for adult content sites and for some features on platforms likely to be used by children. You can usually still create and use social accounts without uploading ID, though you may see more prompts over time.

Is end-to-end encryption illegal in the UK under the Act?

No. Encrypted messaging apps like Signal and WhatsApp remain fully legal. The Act contains a dormant power (Section 121) that could in future require scanning of encrypted content, but the government has confirmed it will only be used when technically possible without undermining encryption — and that threshold has not been met.

Can Ofcom read my private messages?

No. Ofcom regulates platforms, not individual users, and has no power to read your personal messages. Law enforcement can still obtain communications data with appropriate warrants under separate legislation such as the Investigatory Powers Act, as was the case before the Online Safety Act.

What happens if a platform refuses to comply with the Act?

Ofcom can impose fines of up to £18 million or 10% of global turnover, order business disruption measures (such as asking payment processors or ISPs to restrict access), and in extreme cases pursue criminal liability against named senior managers for specific breaches relating to child safety duties.

How can I share links online without exposing my personal data?

Use a privacy-first URL shortener that doesn't sell click data to advertisers, avoid pasting URLs that contain tracking parameters or session tokens, and consider sharing through encrypted channels when the content is sensitive. Tools like Lunyb are designed with these use cases in mind — our 2026 Lunyb review explains the privacy model in detail.

Protect your links with Lunyb

Create secure, trackable short links and QR codes in seconds.

Get Started Free

Related Articles