🧠 Food for thought

Implications, context, and why it matters.

• The Online Safety Act requires services to cut illegal use and remove illegal posts, including intimate image abuse such as non-consensual or AI-generated sexual images 1. X, the social media platform, must also build features that stop this material before it appears 1.
• Illegal content duties started in March 2025, with providers required to finish risk assessments then, while Ofcom consulted guidance on protecting women and girls in February 2025 with final text due later in the year 12.
• Fines can reach £18 million or 10% of qualifying worldwide revenue, with criminal exposure for senior managers in set cases such as failures to meet information requests or certain child safety or child sexual exploitation and abuse (CSEA) orders 2. Ofcom has warned it will act if firms do not protect users 3.
• Vendors of hash‑matching tools, which create fingerprints to spot non‑consensual intimate images, should see uptake as services align with Ofcom expectations 4. Ofcom endorsed this in November 2025 guidance, said adoption is voluntary, and set a 2027 review 4.
• Compliance workflow startups for user‑generated content (UGC) services can plan launches ahead of Ofcom’s 2027 assessment report, as buyers will seek third‑party tools in the 12 to 18 months before the review to prove compliance 4.
• Builders of age assurance and content moderation products should track rules for categorised services, which Ofcom assigns by size and functionality as Category 1, 2A, or 2B 2. Ofcom expected to publish the register in Summer 2025, then consult on extra codes by early 2026, with transparency reports to follow 2.