Check it before the regulator does
Brands and advertisers publish more content on more channels than ever. Metadat-AI helps you find your assets and check them against the rules before they go public.
The challenge
The ASA and CAP now use AI to monitor ads at scale. In 2025 their Active Ad Monitoring system processed nearly 60 million online ads, and 22,383 ads were amended or removed. "Nobody complained" is no longer a safety net.
- Advertising rules are detailed and keep changing, including HFSS (UK) restrictions.
- Small teams have no in-house compliance editor.
- Regulators now use AI to monitor ads at scale, so problems are found without a complaint.
- Brand libraries are huge and hard to search.
Where Metadat-AI helps most:
- Nearly 60 milliononline ads processed by the ASA's AI monitoring in 2025Source: ASA and CAP Annual Report 2025
- 22,383ads amended or removed as a result of ASA and CAP work in 2025
How Metadat-AI helps
- Check against the rules Screen for HFSS (UK) and check against advertising-code categories
- A first pass without extra headcount Risky content is flagged by category and exact moment, so reviewers go straight to it
- Check it before the regulator does Scan your content with AI before it goes public, the way regulators now do
- Find any asset Tag and search your brand library in your own environment
HFSS (UK)
HFSS (high in fat, salt or sugar) restrictions are a UK rule. Since 5 January 2026, ads for identifiable less healthy food and drink products are banned on TV and on-demand services between 5.30am and 9pm, and in paid-for online media at any time. (Source: ASA)
- Find HFSS products on screen Food and drink in your video is identified and checked against the rules
- Flagged by exact moment Reviewers jump straight to the shot, with a timecode, instead of watching every second
- Check the time slot See whether content can air between 5.30am and 9pm
Offensive content
Content that could offend your audience, or break the rules in a market, can damage your brand and attract fines. Metadat-AI identifies offensive content across 50+ categories and classifies ratings (such as BBFC), flagged by category and exact moment.
- Broad categories Alcohol, gambling, drugs, violence, strong language, nudity and more
- Regional rules For example, religious imagery that is inappropriate in some regions
- Your own rules Custom models can check for your brand standards
Models and features used
- Auto-Compliance HFSS screening and advertising-code categories, flagged by moment.About Auto-Compliance
- AI video detection Flags AI-generated or altered footage so ads can be labelled under the EU AI Act.About AI video detection
- Media Tagging Make your brand library searchable.About Media Tagging
- Custom models Check for your own brand rules and products.About custom models
FAQ
Does our media leave our environment?
Not when you run Metadat-AI in your own cloud or on your own hardware.
Does the AI make the final call?
No. AI finds and flags, and your people decide.
Does it work with our existing tools?
Yes. Metadat-AI connects to asset management and creative tools, so your team keeps working where they do today.
Does it replace our compliance sign-off?
No. It flags risks early so that reviewers spend their time on what matters. This is not legal advice.