💻 Pillar 2 · Briefing Card

Online Safety Input — Platform Risks & Officer Value-Add

The Safeguard Hub · safeguard-hub.org/for-police/ · 2025–26 edition · May 2026

What You Bring That Teachers Cannot

  • Operational reality — you've attended cases; young people respond to "I dealt with a case last year" differently from a curriculum resource
  • Criminal framework — the legal consequences of sharing intimate images, possessing CSAM, grooming offences
  • CEOP credibility — explain what actually happens when a report is made and why it's worth doing
  • Perpetrator perspective — what the first grooming message looks like; how escalation works

Positioning Alongside ThinkUKnow (CEOP)

Do not replicate ThinkUKnow — it is professionally designed and free for schools. Position your input as the lived complement: "I'm going to add what actually happens after someone clicks the CEOP report button, and the specific platforms where we're seeing the most concern right now." Recommend ThinkUKnow (thinkuknow.co.uk) to the DSL if the school hasn't used it.

Criminal Consequences to Emphasise

  • Sharing an intimate image without consent — criminal offence under the Online Safety Act 2023. The threat to share (cyberflashing) is also an offence.
  • Possessing CSAM — criminal offence under Protection of Children Act 1978 even if the image was sent unsolicited. Receiving it: report to CEOP, do not forward.
  • "I didn't know they were under 16" — not a defence in law in most circumstances. Juries consider whether reasonable steps were taken to ascertain age.

Current Platform Risks — 2025–26

🎮 Discord
  • No robust age verification
  • Public → private server grooming pathway
  • NSFW channels accessible to under-18s
  • Voice chat creates false intimacy
  • Advise: do not accept server invites from unknown users
🎮 Roblox
  • Primarily under-16 user base
  • Voice chat 13+ but weak age verification
  • Robux = real financial relationship exploited by groomers
  • CEOP cases: grooming began via game invitations
🤖 AI Companion Apps
  • Character.ai, Replika, Chai
  • Young people disclose personal information believing it is private
  • Human actors operating "AI" personas to groom users
  • Online Safety Act 2023 duties apply; Ofcom enforcement developing
  • Advise: anything you tell an AI may be read by a human
🆕 TikTok DMs
  • Algorithm creates rapid intimacy with content
  • DMs can be enabled for under-16s with parental permission — easily bypassed
  • Algorithm surfaces users to potential groomers by interest
  • Unsolicited DM = same risk as any other platform

Closing the Session

End with ceop.police.uk — explain briefly: report goes to NCA, reviewed by investigators, contributes to intelligence even if no immediate action. A report is never wasted. Leave the DSL with ThinkUKnow links for each year group in the session.

Key References & Reporting