Two Different Problems
Bot detection and coordinated inauthentic behaviour detection are related disciplines, but they are not the same thing. Conflating them leads to significant gaps in enterprise narrative defence.
A bot is an automated account. It may or may not be part of a coordinated campaign. A coordinated campaign may or may not use bots. The most sophisticated influence operations today use a combination of automated accounts, real human participants, and AI-generated content in ways that make individual account classification unreliable as the primary detection method.
What Bot Detection Actually Measures
Traditional bot detection classifies individual accounts on a spectrum from automated to human. It uses signals like posting frequency, account age, follower-to-following ratios, and content patterns to make a probabilistic assessment of whether a given account is automated.
This is valuable. Knowing whether an account is real helps assess the organic reach of a conversation. But it does not tell you whether a group of real human accounts is acting in coordination. And it does not tell you whether an automated account is part of an organised campaign or simply a spam operation with no specific target.
What Coordinated Inauthentic Behaviour Detection Measures
Coordinated inauthentic behaviour detection operates at the network level rather than the account level. Instead of asking whether a given account is a bot, it asks whether a group of accounts is behaving in a coordinated way that suggests organised manipulation.
The signals are different. Timing synchronisation across posts. Shared content templates being replicated with minor variations. Amplification sequences where accounts consistently boost each other before any organic engagement. Network structures that reveal a core seeding group and a larger amplification layer. These patterns are visible at the network level even when individual accounts appear entirely human.
This is why some of the most damaging influence operations are not detectable by bot detection alone. The 2020 US election influence operations used large numbers of real human accounts, coordinated by small core networks. The ESG activist campaigns targeting mining and infrastructure companies during shareholder votes are often run by real people using coordinated messaging. Detecting these requires network-level behavioural analysis, not individual account classification.
The Practical Implication for Enterprise Teams
If your narrative monitoring strategy relies solely on bot detection, you are protected against a subset of the threat. Automated spam accounts, simple inauthentic amplification, and obvious bot networks are detectable. But organised human networks using coordinated messaging, activist campaigns using authentic accounts, and sophisticated influence operations using a mix of real and automated participants are not.
A complete enterprise narrative defence requires both layers. Individual account classification to assess the authenticity of individual participants, and network-level behavioural analysis to detect coordination regardless of whether the accounts involved are human or automated.
PulseCheck addresses the first layer. Signal addresses the second. Together they provide the coverage that neither approach delivers alone.
March 16, 2026