How they are detected
How to read a takedown report
Short answer
- Meta definition
- CIB is coordinated efforts to manipulate public debate for a strategic goal, with fake accounts central 1
- Meta focus
- Behaviour, not content 1
- Breakout Scale
- Six categories, from one community on one platform to a policy response or call for violence 6
- Indicators
- Meta publishes threat indicators on GitHub 4
The parts of a Meta network entry
Meta defines coordinated inauthentic behaviour as coordinated efforts to manipulate public debate for a strategic goal, in which fake accounts are central 1. It says it focuses on behaviour, not content, no matter who is behind an operation or whether it is foreign or domestic 1.
Each network summary in the Q1 2024 report uses headings such as "Presence on Facebook and Instagram" and "Followers", and some add ad spend 1. The appendix sorts indicators by stages of Meta's Online Operations Kill Chain, such as acquiring assets, disguising assets and evading detection 1.
Reading follower and size figures
Follower counts measure accounts that followed a Page or account, not people reached. Meta said one network in its Q2 2024 report had about 32,000 Page followers, including inauthentic accounts in Brazil that suggest purchased engagement 2. In the same report Meta urged that these threats be reported with caution 2.
Figures can differ inside one report. Meta's Q1 2025 report gives about 40 Instagram followers for a Romania network in the main text and about 35 in the appendix 3. When two figures differ, cite both with their location.
Impact scores
The Breakout Scale was published by the Brookings Institution in September 2020 as a model for comparing influence operations 6. Category One is spread in one community on one platform. Category Four is breakout to mainstream media. Category Six is a policy response or a call for violence 6.
OpenAI uses the scale in its reports. In May 2024 it said none of five operations scored above Category Two 5.
Google bulletins and EEAS counts
Google's TAG bulletins list coordinated influence operation campaigns terminated on its platforms each quarter, month by month 7. Entries give an origin, such as "Russia" or "a Russian consulting firm", and counts of YouTube channels, Blogger blogs, ad accounts or domains 7. A blocked domain loses eligibility to appear on Google News surfaces and Discover, which is a restriction rather than an account termination 7.
EEAS reports count incidents, not removed accounts. The 3rd report classes channels from state official to state-aligned and says its data is a limited subset that should not be used to draw conclusions about general trends 8.
Threat indicators and attribution
Meta publishes threat indicators, such as domains and account identifiers, in its GitHub repository under the MIT License 4. It says sharing an operation's links or interacting with them would not be enough to attribute accounts to a campaign without corroborating evidence 4,1.
Common questions
Does a takedown mean a crime was committed?
Not on its own. A takedown is enforcement of platform rules. Meta says it removes networks for violating its coordinated inauthentic behaviour policy 1.
Is the follower count the number of real people reached?
No. Meta noted inauthentic followers in at least one network in August 2024 2.
What does Breakout Category Two mean?
Spread in one community across several platforms, or across several communities on one platform 6.
Where are Meta's threat indicators?
In the facebook/threat-research repository on GitHub 4.
Rules and charges change. Check the official source linked on this page before you pay or travel. Spotted an error? Email hello@influenceopsrecords.com.
Sources
- Meta: Q1 2024 Adversarial Threat Report (May 2024)
- Meta: Q2 2024 Adversarial Threat Report (August 2024)
- Meta: Adversarial Threat Report, First Quarter 2025 (May 2025)
- Meta, GitHub facebook/threat-research: Threat Research Indicators
- OpenAI: Disrupting deceptive uses of AI by covert influence operations (30 May 2024)
- Brookings Institution: The Breakout Scale: Measuring the impact of influence operations (September 2020)
- Google: TAG Bulletin: Q3 2024 (12 September 2024, updated 14 January 2025)
- EEAS: 3rd EEAS Report on Foreign Information Manipulation and Interference Threats, Exposing the architecture of FIMI operations (March 2025)
Related
- Meta's adversarial threat reports explained
What Meta's adversarial threat reports contain, how each network entry is laid out, and what the Q4 2024, Q1 2025 and Q2/Q3 2025 reports listed.
- Google TAG bulletins explained
How Google Threat Analysis Group bulletins list terminated YouTube channels, blogs, ads accounts and blocked domains, with 2025 examples.
- OpenAI threat reports explained
What OpenAI's reports on disrupting malicious uses of AI contain, how cases are rated on the Breakout Scale, and what the 2024 to 2026 reports found.
- How attribution of influence operations works
How Meta, Google, OpenAI, Microsoft, X and the EEAS word attribution, from country of origin to named entities, and what terms like likely and linked mean.
- Research datasets on influence operations
Public datasets and archives on influence operations from Meta, Google, OpenAI, the EEAS, Princeton and Oxford, with what each covers and its period.