Why finding it matters
9 in 10
radicalised extremists sentenced in England and Wales in 2019 to 2021 were radicalised at least partly online.
Source: HM Prison and Probation Service, 2022.
Terrorist Content Discovery Index
We measure how easily terrorist content can be found online. Platforms report how much terrorist content they remove. No standard measure shows how easily it can still be found.
The Terrorist Content Discovery Index (TCDI) scores every find on one ten-level scale, on every platform and over time.
The ten levels of discoverability
A lower level means the content was easier to find, which is a more serious failure of detection.
Detection failures
Under 5 minKeyword search
Under 5 minAltered keywords
Under 5 minLogos and flags
Under 5 minKnown propaganda
Moderate evasion
Under 30 minCoded language
1 to 2 hoursLinked accounts
2 to 4 hoursCross-platform links
Effective concealment
1 to 2 daysClosed groups
Several daysDisguised content
Sustained workDeep investigation
Proposed benchmark: basic automation should catch levels 1 to 4. Content found mainly at levels 8 to 10 shows effective detection.
9 in 10
radicalised extremists sentenced in England and Wales in 2019 to 2021 were radicalised at least partly online.
Source: HM Prison and Probation Service, 2022.
Findable
A prevalence figure says how much there is. TCDI measures the effort it takes to find it with public tools.
The index records how each find was made, so it shows where detection falls short, platform by platform.
The five-minute test
The test counts how much terrorist content basic public searches find on a platform within five minutes. It uses only the platform's public search, so it measures what any user could reach.
How it works
Together, the finds show where detection falls short.
Our analysts and trusted flaggers record finds in daily work, and run five-minute tests on major platforms.
Each find gets its level, the time it took, the platform and the content type.
With how long removal takes, the records show how each platform performs over time.
How hard terrorist content is to find on each service.
Whether discoverability is rising or falling.
Where each platform is stronger or weaker than others.
Whether easy-to-find content is removed faster or slower.
Who it serves
Evidence to test platforms' detection claims under UK and EU online safety law.
A shared scale, so countries can compare results and align expectations.
A published method that shows where detection misses content.
One way to record finds, so evidence from many organisations adds up.
What we ask
Regulators
Review the method, then recognise it as an evidence standard in guidance.
Platforms
Review the method before it is published, and tell us where it is wrong.
Funders
Fund regular assessments across major platforms, with public reporting.
Trusted flaggers and researchers
Record your finds on the same ten levels.
Scope and fairness. TCDI covers content from groups on UK, EU, US or UN terrorism designation lists, and material produced by terrorist attackers or glorifying their attacks. The method will be published for review before any results, and platforms are told their results before anything is published.
What comes next
Phase 1
Phase 2
Phase 3
Phase 4
Get in touch
Talk to us about the method, a pilot or a partnership.
Or use the form at techagainstterrorism.org/contact-us.