Decoding the world of cybersecurity

Ofcom proposes wider detection duties for online platforms

Ofcom has submitted tougher online safety codes to Parliament, proposing wider hash matching and proactive detection of child sexual abuse material on specified services from 2027.

Ofcom proposes wider detection duties for online platforms
Summary
  • Ofcom announced stronger measures on 16 November for platforms handling child sexual abuse material.
  • The draft package would expand hash matching and proactive detection for specified services and strengthen default protections against grooming.
  • Some requirements are subject to parliamentary process and expected from early 2027, not immediately enforceable across all platforms.

Ofcom has submitted tougher online safety codes to Parliament, proposing wider use of hash matching and proactive technology to detect and remove child sexual abuse material on specified services. On 16 November, the regulator submitted amendments to existing codes and a separate proactive-technology code under the Online Safety Act. Subject to parliamentary process, the new duties are expected to come into force from early 2027, rather than applying uniformly and immediately to every online service.

The package requires more platforms to use matching technology that recognises known abusive images, including certain pornography services. It also calls for automated tools to identify previously unknown material on specified services, including file-sharing platforms and the largest social networks. Other provisions address accounts that create or distribute abuse material, with requirements for services to take action against offending users.

Hash matching works by comparing a representation of a file against records of previously identified prohibited material. It allows platforms to detect recurring copies or variants without relying solely on human review of every upload. The reliability of a system depends on the reference database, matching method and surrounding review procedures. Ofcom’s measures place greater weight on using those tools in settings where the regulator considers them proportionate to the risk.

Previously unknown material presents a different challenge because it has not yet been added to a known-content reference set. Automated detection can help direct material towards moderation and investigation, but decisions about classification, appeals and evidence handling remain important. The proposed obligations are differentiated by service type and risk. It would be inaccurate to describe the announcement as a universal requirement for all platforms to scan every private communication.

Ofcom also intends to strengthen measures against grooming by ensuring protective settings apply to children’s accounts by default. In some cases, platforms must either conduct highly effective age checks to identify accounts requiring those protections or extend the safeguards to all users. The regulator separately wants services to exercise greater care before recommending content to children, reducing the chance that recommendation systems direct them towards illegal material.

The announcement arrives as the UK prepares for a separate restriction on social media access for under-16s, with implementation criteria expected from government. Ofcom has said that this policy affects decisions about measures previously considered, including specific proposals around children livestreaming. The new announcement does not mean all the measures from earlier consultations will be adopted unchanged; some are being deferred or replaced by more direct regulatory expectations.

The obligations have implications for platform security architecture as well as moderation staffing. Companies may need systems for recognising harmful uploads, retaining trusted reference information, restricting repeat offenders and responding consistently to reports. Such functions sit alongside traditional access control and fraud prevention but address a distinct category of legally prohibited online harm. Their design also raises questions about false positives, oversight and the safeguarding of sensitive evidence.

Ofcom is considering a tougher approach to other forms of illegal content too, with future proposals intended to replace some platform self-assessments with more specific technology requirements. It expects to consult further by spring 2027. That future consultation should not be confused with the measures already announced for child sexual abuse material, nor with existing duties under the Online Safety Act that platforms must already assess and address.

Implementation will depend on the final codes, the parliamentary process and the categories of services covered. The regulator says it expects companies to strengthen detection and removal efforts, while also making child protection measures more systematic. The additional requirements will depend on the parliamentary process and service classifications; Ofcom expects the new measures to take effect in 2027.

×