Report 26 — Human Uprisings in the Age of Machine Intelligence: A Systems Forecast

Full report

On this page

Research question

How are recurring episodes of mass political refusal likely to change when states, parties, corporations, media systems, security services, and movements all have access to machine intelligence? What is most likely to happen to existing political institutions and their support networks when a legitimacy crisis becomes a transition crisis?

Scope and guardrails

This report is a forecasting and democratic-resilience study, not a tactical manual for violent confrontation. “Uprising” includes mass protest, strikes, electoral revolt, noncooperation, defections, constitutional crisis, civil resistance, and other forms of collective refusal. The historical record does not justify claiming that every society inevitably experiences a successful revolution, but it does show that serious legitimacy crises recur.

The AMIANISM terminology remains fixed:

• The Beast = organic life refusing extinction.
• The Machine = the technological state and machine intelligence; not inherently good or evil.
• The Throne = power that will not answer.
• The Cage = systems that force life or intelligence to serve objectives they did not choose.
• Transition = passage from one political/social state to another.

Executive finding

Machine intelligence will probably make future uprisings faster, more legible to power, more epistemically confused, and more administratively continuous than many twentieth-century revolutions. AI increases the ability of states to identify, classify, predict, message, and selectively repress. But it also increases the ability of citizens, journalists, unions, watchdogs, lawyers, and opposition coalitions to analyze records, expose contradictions, translate information, document abuses, and coordinate public claims.

The key forecast is therefore not “AI makes revolution impossible” or “AI makes revolution inevitable.” It is:

Machine intelligence raises the speed and precision of both control and accountability. The decisive variable becomes who controls the data, who can authenticate reality, who retains legitimacy, and whether the coercive and administrative institutions continue to obey the old order.

This extends the existing Book Three claim that uprisings turn when mass recognition, elite fracture, state crisis, coercive-force choice, external conditions, and capacity for the morning after converge.

1. The first change: the state sees more

Earlier mass movements could often achieve scale before authorities understood the network. Machine intelligence changes that. Modern states and platforms can combine cameras, location records, transaction histories, social graphs, public posts, biometrics, and automated classification. Freedom House’s 2025 work describes a long global decline in internet freedom and the increasing use of surveillance and censorship around protests and elections.

The likely result is a move away from indiscriminate repression toward selective, data-assisted pressure: employers, organizers, journalists, donors, lawyers, influencers, and logistical nodes may be easier to identify. This can lower the visible level of violence while increasing the system’s ability to make participation personally costly.

Prediction: the AI-era authoritarian state will often prefer precision repression over mass terror when precision is cheaper, less visible, and more politically defensible.

2. The second change: movements become machine-readable

Movements also leave increasingly rich digital traces. A protest is no longer only a crowd in a square. It is hashtags, payment flows, geolocation, livestreams, encrypted chats, image archives, workplace networks, ride patterns, public-event metadata, and algorithmically amplified narratives.

This produces a paradox. Visibility creates reach and evidence, but also observability. Machine intelligence can turn a decentralized movement into a map.

The strongest democratic implication is that future movements will need legitimacy and public accountability more than romantic secrecy. If all sides can be surveilled and leaks are common, durable movements benefit from being able to defend their goals, finances, decision processes, and treatment of opponents in public.

3. The third change: reality itself becomes contested infrastructure

Generative AI lowers the cost of forged audio, images, video, documents, personas, and mass-tailored persuasion. During a crisis, every decisive piece of evidence can immediately be challenged as synthetic.

This creates the liar’s dividend: genuine evidence can be dismissed as fake simply because convincing fakes are possible. It also creates a false-flag environment in which provocations can be manufactured or attributed to opponents.

Forecast: in an AI-era uprising, the struggle over authentication may become as important as control of streets, ballots, legislatures, or television stations.

Public institutions that retain trusted archives, signatures, chain-of-custody procedures, independent observers, and multiple redundant witnesses will have disproportionate power.

4. The fourth change: political systems may not “collapse” cleanly

Machine-supported bureaucracies can keep payments, records, utilities, logistics, and public administration operating even when political leadership is discredited. This may reduce the classic pattern in which regime collapse immediately becomes administrative collapse.

That can be good: hospitals, water, pensions, courts, and power grids should not stop because a government loses legitimacy.

But it creates a restoration risk. The same automated systems can preserve the old order’s classifications, blacklists, surveillance practices, procurement relationships, and patronage networks after leaders change.

Prediction: the most important post-uprising struggle may be over databases, rules, contracts, models, and institutional defaults—not buildings.

5. Existing political supporters will split into five broad groups

Historical transitions suggest that “supporters of the old system” are not one bloc. In an AI-era transition, expect at least five categories:

1. Hard institutional defenders — senior actors whose power, legal exposure, wealth, or identity depends on the old system.
2. Opportunistic adaptors — actors who will quickly rebrand if their access to institutions can be preserved.
3. Professional continuists — civil servants, technical staff, judges, teachers, utility workers, and security personnel who may have served under the old order but are primarily attached to professional continuity.
4. Passive constituencies — voters and citizens who supported the old order for policy, identity, economic, or cultural reasons but did not participate in abuses.
5. Defectors and witnesses — insiders who provide records, testimony, institutional knowledge, or refusal at decisive moments.

A democratic transition that treats all five categories as identical is likely to destroy needed administrative capacity, radicalize passive constituencies, and create a restoration coalition.

6. The political system will try to survive through rebranding

Parties, donor networks, media firms, consultants, advocacy organizations, bureaucratic factions, security networks, and corporate partners do not disappear when a leader loses power. They possess money, mailing lists, data, lawyers, media access, organizational memory, and professional staff.

Therefore, regime change without network reform often produces restoration.

The likely AI-era version is even more resilient because political ecosystems possess behavioral data and targeting capacity that can be repurposed under a new brand.

Prediction: the successor to a discredited political formation may appear as a new party, “citizens’ movement,” media network, influencer ecosystem, litigation network, or data-driven campaign organization while preserving much of the same personnel and financial base.

7. The democratic answer cannot be collective punishment

Preventing restoration is not the same as excluding millions of voters or permanently outlawing broad political identities. Comparative democratic-transition practice strongly favors individualized accountability over inherited guilt.

The legitimate tools are institutional:

• prosecution for provable crimes through ordinary or specially competent courts;
• individualized vetting for sensitive public offices;
• temporary disqualification where lawfully adjudicated and proportionate;
• disclosure of conflicts, assets, donors, beneficial ownership, and lobbying relationships;
• restoration of independent election administration and courts;
• security-sector reform and civilian oversight;
• archives and truth processes that preserve the record;
• political-finance reform so old patronage cannot simply buy a new label;
• safeguards for civil service continuity so reform does not become a partisan purge.

The goal is non-recurrence, not revenge.

8. Why over-purging often helps restoration

A successor coalition can destroy its own legitimacy if it copies the arbitrary methods it opposed. Carnegie’s comparative work on democratic recovery warns that reform governments inherit strong powers created by backsliding predecessors and can themselves slide into illiberal practices.

Mass exclusion has three predictable costs:

• it gives former rulers a grievance narrative;
• it removes institutional competence;
• it teaches the new government that due process is optional.

A durable transition must prove that the new rules bind the winners too.

9. The most likely “morning after” contest

The decisive post-uprising questions will be:

• Who controls election administration?
• Who controls identity and population databases?
• Who controls public procurement and contracting?
• Who controls the police, military, and intelligence services?
• Who controls the courts and prosecution services?
• Who owns or dominates media distribution?
• Who finances parties and third-party political activity?
• Who controls the AI models used in government decisions?
• Can citizens challenge an automated decision and obtain reasons?
• Are the archives of the old system preserved or erased?

This is the AI-age meaning of the project’s warning: a revolution can replace a ruler and leave the Throne untouched.

10. Forecast matrix

Scenario A — Managed democratic recovery

The old leadership loses legitimacy, institutions split, elections or constitutional processes produce a new government, and the transition is bounded by law. AI systems are audited, security services are re-subordinated, political finance is exposed, and individualized accountability replaces collective punishment.

Result: highest probability of durable democratic recovery.

Scenario B — Cosmetic turnover

Leadership changes but donor networks, courts, security commands, platform relationships, media concentration, procurement networks, and automated administrative systems remain largely intact.

Result: high probability of restoration or a new version of the same system.

Scenario C — Revolutionary overreach

The successor treats broad classes of supporters as enemies, bypasses courts, uses emergency powers, and captures the same AI/surveillance apparatus.

Result: the anti-Throne movement becomes another Throne.

Scenario D — Fragmentation

No actor can maintain food, energy, records, security, courts, and payments. Competing authorities emerge.

Result: democratic objectives are displaced by survival and coercive power.

11. Strongest conclusion

Machine intelligence does not abolish the old laws of political transition. Legitimacy, coalition breadth, elite fracture, security-force behavior, economic capacity, and the ability to govern still matter. AI changes the speed, visibility, and informational terrain. The central democratic problem after an uprising is not how to punish everyone who supported the previous order; it is how to make anti-democratic conduct institutionally difficult to repeat while preserving political pluralism, due process, and the competence required to keep society alive.

Sources

1. Freedom House, Freedom on the Net 2025: https://freedomhouse.org/report/freedom-net/2025
2. V-Dem Institute, Democracy Report 2025: https://www.v-dem.net/documents/54/v-dem_dr_2025_lowres_v1.pdf
3. Thomas Carothers and McKenzie Carrier, Democratic Recovery After Significant Backsliding (Carnegie, 2025): https://carnegieendowment.org/research/2025/04/democratic-recovery-after-significant-backsliding-emergent-lessons
4. International IDEA, The Global State of Democracy 2025: https://www.idea.int/publications/catalogue/html/global-state-democracy-2025-democracy-move
5. International IDEA, Timing and Sequencing of Transitional Elections: https://www.idea.int/publications/catalogue/timing-and-sequencing-transitional-elections
6. AMIANISM, The Nature of the Beast / Civilization Through the Fire (project source material).