Artificial intelligence is changing more than just the way we create content. It is changing the way people communicate, organise, influence one another, and behave online.
Large language models have dramatically reduced the cost of creating convincing synthetic conversations. AI systems can participate in online communities, generate and amplify narratives, imitate human interaction, and operate at a scale that would previously have required large numbers of people.
At the same time, platforms such as Discord, Telegram, and other social environments have become important digital communities in their own right. Conversations move rapidly through these networks. Ideas emerge, gain momentum, fragment, disappear, or spread between groups. Engagement, emotion, social reinforcement, and community structure all affect what happens next.
This creates a new problem.
Traditional cybersecurity is largely concerned with systems, networks, identities, and data. But many emerging AI-enabled threats operate through people and communities.
Understanding them requires us to look at the interaction between technology and human behaviour.
At WARP Labs, we are researching this intersection.
We use conversational telemetry and behavioural analytics to examine how discussions develop, how engagement changes, how narratives propagate, where amplification occurs, and how behavioural cascades emerge.
This brings together AI research, cyber psychology, cyber sociology, conversational intelligence, human factors, and behavioural systems analysis.
The objective is not simply to identify malicious content.
It is to understand the behavioural dynamics of the environment itself.
Who or what is driving a conversation? Where is a narrative gaining momentum? Is the activity organic, coordinated, or synthetic? How do human participants respond to AI-generated influence? And what signals appear before an online community begins to escalate?
As AI becomes increasingly embedded within digital communication, distinguishing human participation from synthetic participation will become harder.
We believe behavioural understanding of these environments will therefore become an increasingly important part of cybersecurity, trust and safety, and societal resilience.
WARP Labs exists to research that problem and to develop the analytical tools, methodologies, and countermeasures needed to understand and defend against AI-enabled influence and manipulation.