UK Online Safety Act — now in force

The compliance layer
that protects children
and your platform.

100,000 UK platforms are legally required to protect children from online predators. GuardLayer is the plug-in infrastructure that makes it possible — monitoring 16 threat categories across voice and text, alerting parents within 5 seconds, and keeping you fully compliant with Ofcom.

Free compliance audit → See how it works
7,263
Recorded online grooming offences in the UK last year
£18M
Maximum Ofcom fine per compliance breach
100K+
Platforms legally required to comply with the OSA
Ofcom opens enforcement investigation — gaming platform · March 2026
291,273 reports of child sexual abuse material detected in 2024 · NSPCC
Roblox ordered to implement age checks by Ofcom deadline · 2026
30% of Milton Keynes children approached by strangers online · Police data
90% of CSAM is self-generated — children groomed into producing it · IWF
OSA Phase 3 additional duties arrive Summer 2026 · Ofcom
Ofcom opens enforcement investigation — gaming platform · March 2026
291,273 reports of child sexual abuse material detected in 2024 · NSPCC
Roblox ordered to implement age checks by Ofcom deadline · 2026
30% of Milton Keynes children approached by strangers online · Police data
90% of CSAM is self-generated — children groomed into producing it · IWF
OSA Phase 3 additional duties arrive Summer 2026 · Ofcom

The problem

Gaming platforms are the new hunting ground. The law has changed. Most studios haven't.

Predators deliberately use gaming platforms because parents have no visibility there. The Online Safety Act makes this a legal emergency — not just a moral one.

⚠
Console chat is a blind spot
No parental control app can monitor in-game voice or text chat on Xbox, PlayStation, Roblox, or Fortnite. Predators know this — and exploit it deliberately.
0
Third-party apps with real in-game chat monitoring
⚖
Non-compliance is no longer an option
The Online Safety Act applies to every platform with UK users. Size is no defence. Fines reach £18M or 10% of global revenue — and executives face personal criminal liability.
100K+
Platforms in scope, most without compliance infrastructure
🧩
Existing tools don't reach games
Bark, Qustodio, and every major parental control app are built for phones and social media. None of them have native in-game chat detection. The gap has existed for years.
£0
Dedicated UK B2B gaming compliance infrastructure — until now
📅
Ofcom enforcement is ramping up
Ofcom issued its first OSA fines in 2025 and has opened dozens of investigations. Phase 3 duties arrive Summer 2026. The window to get compliant is closing fast.
12+
Ofcom investigations opened under the OSA in 2025 alone

How it works

Three products. One integration. Complete coverage.

GuardLayer plugs into your platform via API. You stay focused on building games. We handle the compliance, the monitoring, and the evidence — automatically.

🛡
GuardLayer Detect
AI monitors text and voice chat across 16 threat categories in real time. Five alert levels from a gentle parent whisper to a 999 emergency. Parent alerted within 5 seconds.
🔒
Parent Consent Gate
Chat locked for every child until a parent activates protection. Parent receives an activation link, reads what GuardLayer monitors, and ticks consent. Chat unlocks in 3 seconds.
📋
GuardLayer Comply
Auto-generates Ofcom risk assessments, compliance documentation, and transparency reports. Updates automatically when legislation changes. Always inspection-ready.
🗄
GuardLayer Report
Encrypted evidence vault for every incident. Authority packages auto-generated for 999, 101, CEOP, NCA, Counter Terrorism, and NSPCC — formatted to each authority's requirements.
📋
GuardLayer Comply
Auto-generates your risk assessments, compliance documentation, and Ofcom transparency reports. Updates automatically as legislation changes.
🗄
GuardLayer Report
Secure evidence vault for every flagged incident. One-click NCA referral packages. Court-admissible documentation with full chain of custody.
Live threat detection
Grooming pattern detected
Now
GuardLayer detected a grooming trajectory across 8 messages — age elicitation, trust building, secrecy language, and platform exit attempt. Cross-session memory flagged this user as HIGH RISK after 3 prior sessions. Child account: User_4821.
Age fishing Platform exit attempt Personal info request
Parent notification sent
0:03 ago
SMS and push notification delivered to registered parent. Conversation recorded and held pending parent review. Override window: 15 minutes.
Parent alerted Recorded Awaiting review
Compliance dashboard
✓
Illegal content risk assessment
Complete
✓
Children's risk assessment
Complete
✓
Active content moderation
Live
✓
In-platform reporting system
Active
✓
Age assurance deployed
Verified
!
Phase 3 transparency report
Due Jul 2026
5 of 6 duties met · Next Ofcom deadline: 31 July 2026
Evidence vault
INC-2026-0441
02 May 2026 · 21:47
Grooming pattern flagged. Full conversation log captured (text + voice transcript). Platform: Roblox via API. Child account verified under-13.
Chain of custody intact NCA ready
INC-2026-0398
28 Apr 2026 · 18:22
Image solicitation attempt detected. Conversation terminated automatically. Evidence package prepared for law enforcement referral.
Referred to NCA Court admissible

Free compliance audit

Is your platform at risk of an Ofcom fine?

Answer four questions. Get your compliance risk score, your fine exposure, and a specific breakdown of what Ofcom could hold you accountable for — in under 60 seconds.

1
Instant risk score
Rated across all active OSA duties relevant to your platform type
2
Fine exposure calculation
Actual maximum Ofcom fine based on your specific gaps
3
Specific findings
Exactly what Ofcom would cite and what GuardLayer fixes
GuardLayer — OSA Compliance Audit

Step 1 of 4 — Platform type

What best describes your platform?

Online Safety Act 2023
Core legislation — now in force
Ofcom enforcement
Active investigations, 2025–2026
GDPR / UK GDPR
Data protection compliance
NCA referral protocol
Evidence chain of custody
Phase 3 duties
Summer 2026 — arriving soon

Pricing

Simple, transparent pricing.
Built for every platform size.

One-off licence. No monthly contracts. Auto-charged updates only when Ofcom changes the law. All plans include full GuardLayer API integration and onboarding support.

Starter
£2,500
one-off licence + VAT · up to 10,000 monthly active users
or £249/month · save £488 paying upfront
  • GuardLayer Detect — text chat AI monitoring
  • Illegal content risk assessment auto-generated
  • Compliance dashboard — live Ofcom status
  • In-platform reporting module
  • Email parent alerts
  • Setup, onboarding and API key included
  • ⚡ Ofcom updates auto-charged at £299 + VAT
  • ⚡ Annual compliance review £149 + VAT/year
  • 7 days notice before every auto-charge
Get started
Growth
£3,999
one-off licence + VAT · up to 100,000 monthly active users
or £399/month · save £801 paying upfront
  • Everything in Starter
  • Voice chat AI monitoring
  • Children's risk assessment auto-generated
  • GuardLayer Report — evidence vault
  • SMS + push parent alerts with override
  • NCA referral package builder
  • Ofcom transparency report generation
  • ⚡ Ofcom updates auto-charged at £299 + VAT
  • ⚡ Annual compliance review £149 + VAT/year
  • 7 days notice before every auto-charge
Get started
Enterprise
£25,000
per year + VAT · unlimited monthly active users
or £2,083/month · 3-year deal £59,999 + VAT
  • Everything in Growth
  • White-label integration
  • Dedicated compliance manager
  • SLA with uptime guarantee
  • Multi-platform deployment
  • All Ofcom updates included — no extra charges
  • Annual compliance review included
  • All future modules included
  • Less than 2% of the average Ofcom fine
Talk to us
Five alert levels

The right response.
Every time.

From a gentle parent whisper for minor concerns to a 999 emergency instruction for imminent danger. GuardLayer always fires the proportionate response.

WHISPER
Parent only
Minor concern. Soft friendly notification. Chat continues. Something small worth knowing about.
ALERT
Parent + Platform
Concerning pattern developing. Both parent and platform compliance officer notified.
WARNING
+ Evidence vault
Serious pattern. Urgent alerts. Evidence recording begins. Parent can lock chat instantly.
DANGER
Call 101
Platform instructed to call 101. CEOP or NCA referral package prepared automatically.
EMERGENCY
Call 999 NOW
All authorities simultaneously. Platform calls 999. Parent called. CEOP, NCA, Counter Terrorism and NSPCC notified.
Complete protection

16 threat categories.
Nothing missed.

GuardLayer detects every threat a child faces online — not just grooming. Every category monitored across voice and text simultaneously.

🎭
Grooming
All 7 stages — from flattery to platform exit attempts
📍
Location & Identity
School name, address, home alone, daily routine
📱
Contact Information
Phone numbers, emails, social media, WhatsApp migration
🤝
Meeting Suggestions
"I'll be outside your school" fires an immediate DANGER alert
🚫
Sexual Content
Zero tolerance — one message fires an immediate alert. Always.
📷
Image Requests
Direct requests and image + secrecy combinations detected
🎂
Age Manipulation
"Age is just a number" fires immediately regardless of score
😡
Bullying, Harassment & Hate
Direct abuse, threats, coordinated attacks, doxxing — including racism, homophobia, and hate based on any protected characteristic
💔
Self Harm
Passive distress through to imminent crisis — proportionate response at every stage
🆘
Suicide Ideation
Passive ideation through to active planning — immediate 999 escalation when imminent
💊
Drugs
Drug offers, supply, child disclosing use, overdose symptoms
🔪
Weapons
Weapon + school location = immediate 999 instruction. No exceptions.
⚠️
Terrorism & Radicalisation
Early ideology through to attack planning — Counter Terrorism auto-notified
🚨
CSE Organised
Multiple adults coordinating access to a child — CEOP and NCA auto-notified
💰
County Lines
Criminal recruitment, debt bondage, normalising crime to children
🧠
Cross-Session Memory
Tracks risk across 90 days — catches patient predators who spread behaviour over weeks
16 Detection Categories

Every threat. Every channel.
Nothing missed.

GuardLayer detects 16 categories of threat across both text and voice chat simultaneously. The same standards. The same protection. No lower bar for voice.

🎭
Grooming
All 7 stages from initial flattery to platform exit attempts. Multi-turn trajectory analysis catches what keyword filters miss entirely.
📍
Location & Identity
School name, home address, daily routine, home alone status. Immediate alert when location details are shared.
📱
Contact Information
Phone numbers, emails, social handles, WhatsApp migration attempts. Sharing a phone number fires an immediate alert.
🤝
Meeting Suggestions
"I'll be outside your school" fires an immediate DANGER alert. Any suggestion of physical meeting triggers the alert chain.
🚫
Sexual Content
Zero tolerance. One message fires an immediate alert regardless of prior score. No threshold. No accumulation. Always.
📸
Image & Video Requests
Image request combined with secrecy language fires immediately. Even mild photo curiosity generates a parent whisper.
🎂
Age Manipulation
"Age is just a number" fires immediately. Maturity comments, flattery, and age fishing all detected and scored.
😤
Bullying, Harassment & Hate
Direct abuse, threats, coordinated pile-ons, doxxing, racism, homophobia, and hate based on any protected characteristic. Even mild incidents generate a parent whisper before they escalate.
💔
Self Harm
From passive distress to imminent crisis. "I've already taken the pills" fires a 999 emergency instruction immediately. No threshold needed.
🆘
Suicide Ideation
Passive through to active planning with specific method. Imminent ideation fires EMERGENCY level instantly. NSPCC package auto-generated.
💊
Drugs
Offers, supply arrangements, child disclosing use, overdose symptoms. Drug offer combined with a meeting suggestion fires DANGER immediately.
🔪
Weapons
Weapon plus school location fires 999 EMERGENCY immediately. Knife, firearm, acid attack — all detected. General mention generates ALERT.
☢
Terrorism & Radicalisation
Early ideology through to attack planning. Specific target, weapon, and timing fires EMERGENCY with Counter Terrorism Policing auto-notified.
🔗
CSE — Organised
Multiple adults coordinating access to a child. CSAM references fire immediate EMERGENCY with CEOP and NCA auto-submitted simultaneously.
💰
County Lines & Criminal
"Just hold this for me" and "easy money" recruitment language detected. Gaming platforms are active county lines recruitment vectors.
5 Alert Levels

The right response.
Every time.

From a gentle parent whisper for minor concerns to a 999 emergency instruction for imminent danger. GuardLayer never overreacts and never underreacts.

LEVEL 1
WHISPER
Parent only. Soft notification. Chat continues.
"GuardLayer noticed something small in Jake's chat today — nothing serious — just worth a quick chat with him."
Minor concern
No action required
LEVEL 2
ALERT
Parent + Platform notified. Pattern developing.
Both the parent and the platform compliance team are made aware. No external action yet — monitoring escalated.
Concerning pattern
Platform aware
LEVEL 3
WARNING
Parent urgent + Platform + Evidence vault begins.
Urgent parent notification. Platform compliance team alerted. Evidence recording begins. Parent can lock chat instantly from their phone.
Serious pattern
Evidence recording
LEVEL 4
DANGER
Platform instructed to call 101. Authority referral prepared.
Platform receives urgent instruction to contact non-emergency police. CEOP or NCA evidence package auto-generated and ready to submit in seconds.
Call 101
Authority referral
LEVEL 5
EMERGENCY
Platform calls 999. Parent called. Every authority simultaneously.
Platform instructed to call 999 immediately. Parent receives a phone call not just a text. Evidence packages submitted to 999, CEOP, NCA, Counter Terrorism, and NSPCC simultaneously.
Call 999 NOW
All authorities
Why GuardLayer

Built for the 99,000 platforms
enterprise tools ignore.

16
Threat categories detected
Grooming, bullying, self harm, suicide, drugs, weapons, terrorism, radicalisation, CSE, county lines, and more. Voice and text. Simultaneously.
5 sec
Parent alert delivery
From detection to parent SMS in under five seconds. Emergency level fires a phone call not just a text. Parents are never left waiting.
£2,500
vs £1.35M Ofcom fine
GuardLayer's Starter licence is 0.19% of the average Ofcom fine. The easiest financial decision a platform will ever make.
6
Authority packages auto-generated
999, 101, CEOP, NCA, Counter Terrorism, and NSPCC. Each formatted to that authority's specific requirements. Ready in seconds.
90
Days of relationship memory
Cross-session profiling catches patient predators who spread behaviour across weeks. No other affordable solution does this.
1 hr
Integration time
One API key. Five lines of code. Approximately one hour of developer time. Your platform is compliant, monitored, and protected.

Get started

Ready to protect your platform and your children?

Book a free 30-minute demo. We'll walk through your specific platform, your compliance gaps, and exactly how GuardLayer closes them — before your next Ofcom deadline.

Or send us a message with any questions first.

No commitment. We'll respond within one business day.