Date: 27 February 2026
Speaker: Stephen H. Campbell
Chief Technology Officer for the DISARM Foundation, founder of Non-State Threat Intelligence, and advisor to eosedge Legal.
Stephen H. Campbell used the public disorder in Southport in 2024 as a case study to argue that disinformation, amplified by algorithmic incentives and artificial intelligence, has become a primary accelerant and catalytic driver of civil unrest. He contended that this development necessitates the establishment of national early-warning systems, behavioural intelligence frameworks, and stronger regulation of digital platforms.
You can read a summary of the key points below or download our readout for a full list of definitions. Stephen has also kindly provided a video recording of his session for those that couldn't make it.
Key Points & Takeaways
The Information Environment Has Fundamentally Changed
- Traditional media gatekeeping (editors, anchors, institutions) has, to a large extent, been replaced by algorithms, influencers, and Big Tech platform owners.
- Engagement-driven algorithms maximise engagement by playing to confirmation biases and prioritising emotion, outrage, and polarisation, not truth.
- We now operate in an era of narrative dominance, where facts matter only insofar as they support narratives and disinformation spreads faster and more effectively than correction.
- AI is dramatically lowering the cost, speed, and scale of disinformation, worsening these dynamics.
#1: The current information ecosystem structurally amplifies emotionally engaging and polarising narratives that provoke outrage.
Disinformation Acts as an Accelerant for Real-World Violence
- The Southport incident demonstrates how a real-world tragedy combined with false attribution rapidly escalated into mass mobilisation and violence.
- Disinformation transforms local incidents into national unrest.
- This follows a consistent pattern: Grievance→ Hate → Disinformation → Trigger Event → Mobilisation → Violence
#2: Disinformation is not merely misleading; it can act as a catalytic accelerant, contributing to civil unrest and public disorder.
Economic Incentives Drive Disinformation Spread
- Much of the amplification came from monetised clickbait websites, not state actors.
- Outrage, fear, and polarisation generate traffic and advertising revenue.
Platform algorithms structurally reward this behaviour.
#3: Disinformation is financially incentivised, meaning systemic harm is embedded within prevailing digital platform business models.
Intelligence and Early Warning Systems Failed
- Strategic indicators of rising hate and grievance were visible prior to Southport.
- Data from civil society groups showed rising hate incidents and increasing polarisation.
- Yet public disorder risk was assessed as low.
#4: The intelligence failure was systemic, not operational.
Fusion Centre (Central Monitoring System) Is Needed
- There is no central system/body in the UK responsible for continuous monitoring of disinformation threats, integrating data streams, and producing national early warnings.
#5: The UK requires a permanent national-level monitoring and fusion centre for disinformation, hate escalation, mobilisation risk, and public disorder forecasting.
Behavioural Analysis Is More Effective Than Content Moderation
- Focusing on behaviours avoids free speech debates and regulatory paralysis.
- DISARM provides a behaviour-based taxonomy and a standardised way to describe and track manipulative information operations.
- This enables shared situational awareness, cross-agency intelligence sharing, and predictive modelling based on behavioural fingerprints.
#6: Tracking behaviours, not content, is the most scalable, legally robust, and operationally viable intervention strategy.
Disinformation Risk Can Be Quantified
- A disinformation impact scale can measure engagement, virality, cross-platform spread, influencer amplification, and calls to action
- Escalation thresholds could trigger police response, platform takedowns, and emergency coordination.
#7: Some dimensions of disinformation and mobilisation can be operationalised into measurable indicators and thresholds for early warning purposes.
Platforms Are Risk Multipliers
- Platforms such as X (formerly Twitter) act as algorithmic hate amplifiers and mobilisation accelerators.
- Current regulation remains insufficient.
#8: Without strong platform regulation and algorithmic accountability, civil unrest driven by disinformation could continue to scale.
Lawful-but-Awful Content Remains a Major Gap
- Much harmful content does not meet illegality thresholds yet contributes significantly to radicalisation and mobilisation.
#9: Regulatory frameworks should address lawful but harmful content, not only illegal speech.