The Online Safety Act has moved UK messaging platforms from a reactive posture to a proactive one. As of January 2026, Priority Offences require automated scanning to intercept harmful content before it reaches a recipient — not after a user reports it.
Cyberflashing, serious self-harm, and CSEA are now statutory screening categories requiring automated detection.
Platforms must route detected CSAM signals directly to the National Crime Agency via the new statutory reporting portal.
Ofcom publishes binding requirements on hash-based detection for non-consensual intimate images (NCII).
Category 1 and 2A platforms face doubled transparency reporting obligations once user thresholds are confirmed.
Failure to comply triggers an immediate Category 1 risk assessment from Ofcom. Fines reach up to 10% of annual global turnover.
ProntoScreen covers all statutory screening mandates in a single integration, with a unified verdict and audit trail per message.
Analyses images in outbound private messages for unsolicited intimate content before delivery. Supports still images and video frame sampling.
Detects Child Sexual Exploitation and Abuse material with NCA-reportable output. Generates the structured report payload required by the April 2026 statutory portal.
Flags images and text that encourage or instruct self-injury, giving platforms the signal to intercept delivery and route to safeguarding workflows.
Evaluates message text for grooming patterns, coercive language, solicitation signals, and hate speech — across public channels and private DMs.
From 2026, platforms handling AI-generated media must verify watermark presence. ProntoScreen detects C2PA-compliant watermarks and flags unattributed synthetic content.
For end-to-end encrypted platforms, ProntoScreen offers an on-device screening SDK. Content is screened locally before encryption — satisfying OSA mandates without breaking E2EE.
ProntoScreen sits in your message pipeline as a lightweight interception layer. Every outbound message passes through the API before encryption and delivery. The verdict is returned in under 200 ms.
Your platform passes the outbound message payload — image, text, or both — to the ProntoScreen API endpoint before it is encrypted and sent.
The image and text components are screened simultaneously across all active detection categories. Analysis typically completes in 120–180 ms.
A structured JSON verdict is returned: PASS, FLAG, or BLOCK — with per-category confidence scores, an audit ID, and any NCA report payload where required.
Route to your delivery pipeline (PASS), hold for human review (FLAG), or drop the message and notify the sender (BLOCK). Webhooks available for async flows.
The central tension in OSA compliance is reconciling screening mandates with end-to-end encryption. ProntoScreen offers two deployment modes so you can satisfy both.
For platforms without E2EE, or where temporary message access is permissible under your privacy policy. The message payload is sent to ProntoScreen before it enters the encryption layer. No content is retained post-analysis.
For fully E2EE applications. The ProntoScreen SDK runs on-device, screening content locally against hash databases before the message is encrypted. Nothing leaves the device unencrypted — E2EE remains intact.
Zero-retention by default. ProntoScreen does not store message content, images, or user identifiers after a verdict is issued. Analysis is stateless. Audit logs contain only the verdict, confidence scores, and a pseudonymous session ID — never the original content.
⚠ Yellow indicates partial support requiring additional human moderation workflows.
DMs, group chats, and content feeds. Screen images and text across all message types with a single API integration.
Mandatory cyberflashing detection for unsolicited image sharing. OSA compliance for platforms where intimate image sending is a primary risk vector.
Real-time text moderation for in-game chat and live comment streams, with configurable thresholds per content category and audience age group.
Scan supplier and buyer communications for compliance. Protect users from unsolicited content and detect grooming patterns in business communications.
"We had six weeks to achieve OSA compliance before Ofcom's enforcement window opened. ProntoScreen gave us pre-delivery screening, audit logs, and NCA-ready reporting in under two weeks of integration."
"The client-side SDK solved a problem we'd been stuck on for months — how to screen content without breaking our E2EE promise to users. It's the only solution we found that genuinely did both."
"Having a single API endpoint return a structured verdict with per-category confidence scores, a pseudonymous audit ID, and the NCA payload — ready to submit — removed months of compliance engineering from our roadmap."
Meet the Highly Effective Age Assurance standard using facial estimation or ID document verification with liveness detection.
Learn MoreIdentity-verify users in peer-to-peer contexts — dating, social, and marketplace platforms — before they can initiate private messaging.
Learn MoreAttribute and watermark AI-generated media before it enters your platform — meeting the 2026 synthetic content disclosure requirements.
Learn MoreGet ProntoScreen integrated before Ofcom's next enforcement window. CSEA reporting to the NCA is now mandatory — and the architecture to support it takes less than a day to deploy.
Zero retention · Sub-200 ms latency · NCA reporting built in