Soft Radicalisation: How Civilisations Harden Before They Break. Part 4 — AI Is Not a Religion. It Is an Accelerant.

Every civilisation that slid into violence believed the same thing at the critical moment: “This time is different.”

Different threats. Different tools. Different circumstances.

They were wrong.

Artificial intelligence does not change the nature of radicalisation.

It changes its speed, scale, and deniability.

AI does not create hatred.

It compresses the distance between moral hardening and moral action.

The Wrong Question Everyone Is Asking

Public debate around AI is dominated by the wrong anxieties:

Will AI become conscious?

Will AI replace humans?

Will AI become a god?

These questions are dramatic and largely irrelevant to radicalisation.

The real question is quieter and far more dangerous:

What happens when humans outsource judgment at civilisational scale?

Because radicalisation does not require belief in a god.

It requires obedience without moral friction.

AI Is Not a Belief System — It Is a Decision System

Religion tells people why to live.

AI increasingly tells people what to do next.

That difference matters.

AI does not offer transcendence, salvation, or meaning.

It offers:

Recommendations

Rankings

Risk scores

Predictions

Optimised choices

And it does so with an authority that feels neutral, objective, and impersonal.

This is not theology.

This is procedural authority.

The Core Shift: From Moral Judgment to Procedural Obedience

Soft radicalisation accelerates when judgment is replaced by process.

Historically:

People deferred to priests

Then to ideology

Then to the state

Now, increasingly, they defer to systems.

When a system says:

“This group is high-risk”

“This content is dangerous”

“This outcome is optimal”

“This action reduces threat”

Human responsibility quietly retreats.

Not because people are evil but because delegation feels safer than judgment.

Why AI Compresses the Radicalisation Timeline

AI does not invent new moral failures.

It removes the pauses that once slowed them down.

Here is what changes:

👉 Decisions become faster

👉 Exceptions become automated

👉 Responsibility becomes distributed

👉 Accountability becomes unclear

Historically, moral hardening took decades.

With AI, it can take months—or weeks.

That compression is unprecedented.

⚠️ AI as an Institutional Force Multiplier

The real danger is not individuals using AI.

It is institutions embedding AI into governance, security, moderation, and policy.

When AI is integrated into:

Policing

Surveillance

Risk assessment

Content moderation

Border control

Welfare eligibility

Soft radicalisation gains bureaucratic efficiency.

Exclusion no longer feels personal.

It feels procedural

The Historical Parallel We Should Not Ignore

Consider the Inquisition again not as cruelty, but as process.

The most dangerous feature of the Inquisition was not fanaticism.

It was institutionalised suspicion, processed through rules.

AI introduces a modern equivalent:

Suspicion at scale

Risk without explanation

Judgment without moral context

This is how moral violence becomes administrative.

Why AI Feels “Safer” Than Human Judgment

Humans are uncomfortable with:

Moral ambiguity

Responsibility

Being wrong

AI offers relief:

Clear outputs

Statistical justification

Distance from consequences

This creates a psychological trap.

When outcomes are harmful, people say:

“The system recommended it.”

Soft radicalisation thrives on exactly this abdication.

Important Clarification: Violence Is Not Inevitable

Let’s be precise.

AI does not make violence inevitable.

But without conscious human intervention, it makes radicalisation predictable.

The danger is not AI itself.

It is unquestioned reliance on AI within already hardened moral environments.

AI accelerates whatever values are already embedded.

If those values are exclusionary, fearful, or absolutist—AI will amplify them efficiently.

Ethical AI Will Not Save Us

This is the uncomfortable part.

Ethical AI frameworks focus on:

Bias reduction

Transparency

Fairness

Accountability

These are necessary but insufficient.

Why?

Because soft radicalisation does not require unfair systems.

It requires morally justified exclusion.

A system can be:

Transparent

Fair by defined metrics

Legally compliant

And still participate in radicalisation if the moral framing is already hardened.

Technology cannot compensate for moral abdication.

The Real Risk: Authority Without Ownership

The most dangerous shift AI introduces is this:

Authority increases while ownership of moral responsibility decreases.

When no one feels responsible:

Violence feels accidental

Harm feels systemic

Accountability evaporates

This is the perfect environment for radicalisation to move from thought to action without anyone feeling guilty.

Why This Moment Matters

For the first time in history, societies are deploying systems that:

Decide faster than humans

Scale beyond human comprehension

Shape perception, trust, and fear

Operate without moral intuition

This is not dystopia.

It is governance reality.

The question is not whether AI will be used.

It already is.

The question is whether human judgment will remain sovereign.

🧨 Final Provocation

Radicalisation does not need AI to exist.

But AI makes it efficient, quiet, and deniable.

The future danger is not machines turning against humans.

It is humans hiding behind machines.

In the final part of this series, we confront the hardest question of all: if technology will not save us, what kind of leadership might?

A blog by RK Vedant