The Breakout Scale: How to Measure the Impact of Influence Operations
In short: The Breakout Scale is a six-category scale for judging the potential impact of an influence operation by how far its content spreads: from one community on one platform (Category 1) to mainstream media (4), celebrity amplification (5) and a policy response or call for violence (6). Ben Nimmo proposed it in a Brookings paper in September 2020, and platforms such as OpenAI use it in their threat reports.
Learn to assess influence operations: start FIMI 101 → debunkacademy.org Get new FIMI analysis by email → debunk.org/subscribe
Also known as: Nimmo's Breakout Scale; the breakout scale for influence operations.
Key facts
The Breakout Scale divides influence operations into six categories according to how many platforms and communities their content reaches and whether it breaks out into media, celebrity amplification or policy.
Ben Nimmo proposed the Breakout Scale in The Breakout Scale: Measuring the Impact of Influence Operations, published by Brookings in September 2020.
The Breakout Scale measures potential impact from observable evidence, not actual effect: Nimmo notes that a breakout "does not necessarily equate to actual impact".
An operation can move up and down the Breakout Scale over time; each rating describes the operation at a specific moment.
In May 2024, OpenAI rated five covert influence operations, including Russia's Doppelganger, on the Breakout Scale: none scored higher than Category 2.
Debunk.org analysis of Operation Overload emails received, September 2023 – July 2026: 1,166 emails asked Debunk.org to "verify" fabricated content, an attempt to get fact-checkers to carry the fakes to new audiences.
Updated October 2026 · By Debunk.org, an independent non-profit that researches disinformation
Contents: What is the Breakout Scale? · Who created it? · The six categories · Key ideas: insertion points and breakout moments · Examples · How to use it · Limits · Breakout Scale, ABCDE and DISARM · FAQ · Sources
What is the Breakout Scale?
The Breakout Scale is a comparative tool that rates an influence operation from Category 1 to Category 6 by how far its content has spread beyond the place where it was planted.
Measuring the impact of disinformation is hard. Likes and views can be faked, and the people running an operation rarely say what they wanted to achieve. Ben Nimmo's answer, in a 2020 paper for Brookings, was a scale "based on data that are observable, replicable, verifiable, and available from the moment they were posted". [1]
Nimmo wrote the paper as Head of Investigations at Graphika; he had earlier worked at the Atlantic Council's DFRLab. [1] The scale is "actor-agnostic": it can be applied to state operations, conspiracy theories, health misinformation or any other online campaign. [1] Debunk.org uses it in its own analysis, as described on our methodology page.
Who created the Breakout Scale? Ben Nimmo
Ben Nimmo created the Breakout Scale, in a paper published by Brookings in September 2020, written while he was Head of Investigations at Graphika.
Nimmo, who had earlier worked at the Atlantic Council's DFRLab, designed the scale for "operational researchers" who expose influence operations in real time, including teams at the platforms. [1] Platforms have since used it in their threat reports; OpenAI, for example, rated the operations in its May 2024 report on it. [2]
The six categories
The six Breakout Scale categories run from one community on one platform (1) to policy response or a call for violence (6).
Category | Name | What it means |
1 | One platform, no breakout | The content stays within one community on one platform |
2 | One platform with breakout, or many platforms without breakout | It reaches several communities on one platform, or appears on several platforms but stays in one community |
3 | Multiple platforms, multiple breakouts | It spreads across several platforms and reaches several communities |
4 | Cross-medium breakout | It leaves social media: mainstream media report or repeat it |
5 | Celebrity amplification | High-profile people, such as celebrities or political candidates, amplify it |
6 | Policy response or call for violence | It triggers a policy response or other concrete action, or includes a call for violence |
Source: Nimmo (2020), executive summary and pp. 5–9. [1]
Nimmo calls Category 3 "a transient" stage: operations tend either to fall back or to move on into mainstream media. [1]
Get new FIMI analysis by email. debunk.org/subscribe
Next: learn to rate and document real influence operations in the FIMI 101 course.
Key ideas: insertion points and breakout moments
An operation's "insertion point" is where it first plants content; a "breakout moment" is when real users carry that content into a new community, platform or medium.
Nimmo compares a planted story to a virus inserted into an ecosystem: one infected account matters little, but a story that jumps from community to community, platform to platform and into "real-life discourse" is the one to worry about. [1] Breakout moments can happen:
on-platform: a genuine user repeats the message to a new audience;
cross-platform: someone screenshots it and posts it on another platform;
cross-medium: traditional media embed or report it;
through influencers: a politician or celebrity repeats it;
into policy: decision-makers act on it. [1]
Each breakout brings the content to a larger or more influential audience, which is why the categories form a rising scale of potential impact. [1]
Examples
Most influence operations exposed by researchers and platforms stay in the lowest Breakout Scale categories.
Operation | Rating | Who rated it, when |
Five operations including Doppelganger (Russia), Bad Grammar (Russia) and Spamouflage (China) | None higher than Category 2; Bad Grammar stayed at Category 1 | OpenAI, May 2024 [2] |
Russia's Internet Research Agency (IRA) | Category 6 in May 2016, when it organised two opposing demonstrations in Houston, Texas; lower at other times | Nimmo (2020) [1] |
Russian hack and leak of the US Democratic National Committee, 2016 | Category 6: the leaks led to the resignation of senior DNC staff | Nimmo (2020) [1] |
OpenAI described Category 2 as "posting activity on multiple platforms, but with no breakout or significant audience engagement in any of them", and warned that operations "could … reach the higher categories in the future, especially if they are unwittingly amplified by media, influencers or politicians". [2]
What this means for fact-checkers. Nimmo warned that journalists "risk bringing an IO to new audiences if they fall for its messages". [1] Operation Overload is built around that risk. Between September 2023 and July 2026, Debunk.org received 1,166 of its emails asking us to "verify" fabricated, media-branded content. [3] Read through the Breakout Scale, every such email is an attempt at a cross-medium breakout (Category 4) that the operation can't achieve on its own: a debunk that reproduces the fake, without a watermark or context, would carry it to an audience it never reached. That is why Debunk.org describes such fakes without amplifying them, and why the EEAS and ISD found that Operation Overload itself reaches few real audiences. [3] (Debunk.org's reading of the scale, not a rating by Nimmo.)
How to use the Breakout Scale: 5 steps
To rate an operation, find where it started, track where it spread, and assign the highest category supported by evidence at that moment.
Find the insertion point(s): where did the content first appear, and in which community?
Look for breakouts: did unaffiliated, genuine users repeat it? On other platforms? In mainstream media? Did a public figure share it?
Separate the operation from its amplifiers: reposts by the operation's own fake accounts are not a breakout.
Assign the category supported by the evidence, and record the date: the rating is a snapshot.
Re-rate over time: operations can rise and fall on the scale.
Limits
The Breakout Scale shows how far content travelled, not whether it changed minds or votes.
Potential, not actual impact: a politician may repeat a claim and see it rejected. [1]
Intent unknown: without knowing what the operators wanted, researchers can't say whether an operation "succeeded". [1]
Visibility: private groups and messaging apps hide breakouts from outside researchers.
Reporting can raise the rating: media coverage of an operation can itself be a breakout if it repeats the content uncritically; careful reporting describes the operation without spreading its message.
Breakout Scale, ABCDE and DISARM
The Breakout Scale answers "how far did it spread?", which corresponds to the Degree component of the ABCDE framework; DISARM describes the techniques used to try to spread it.
Framework | Question | Debunk.org guide |
Breakout Scale | How far did it break out (1–6)? | this page |
ABCDE | Actor, behaviour, content, degree, effect | |
DISARM | Which techniques were used? |
All three are tools for analysing foreign information manipulation and interference, explained on our What is FIMI? page.
Practise rating real operations. FIMI 101 teaches the frameworks analysts use to assess influence operations, with real cases. → Start FIMI 101
FAQ
What is the Breakout Scale? A six-category scale, proposed by Ben Nimmo in 2020, that rates influence operations by how far their content spreads across communities, platforms, media and policy.
What is Category 6 on the Breakout Scale? An operation that triggers a policy response or other concrete action, or that includes a call for violence.
Does a high Breakout Scale category mean an operation worked? Not necessarily. The scale measures potential impact from observable spread; actual effects need separate evidence.
Who uses the Breakout Scale? Researchers and platforms; for example, OpenAI rated the influence operations in its May 2024 threat report on it.
Written by Debunk.org. Updated October 2026. Debunk.org is an independent non-profit that researches disinformation and FIMI and an EU Digital Services Act trusted flagger.
Sources
Ben Nimmo, The Breakout Scale: Measuring the Impact of Influence Operations, Brookings (Foreign Policy), September 2020. https://www.brookings.edu/wp-content/uploads/2020/09/Nimmo_influence_operations_PDF.pdf (accessed 23 Oct 2026)
OpenAI, AI and Covert Influence Operations: Latest Trends, May 2024. https://openai.com/index/disrupting-deceptive-uses-of-ai-by-covert-influence-operations/ (report text read from a saved copy, 23 Oct 2026)
Debunk.org, Operation Overload profile (Debunk.org email analysis; EEAS 4th FIMI Threat Report, March 2026; ISD, July 2025); Debunk.org, Operation Overload: Debunk.org's case. https://www.debunk.org/operation-overload-debunk-org-s-case
Comments