When Electricity Becomes Governance
Why Energy Regulators Are Entering the AI Era
Author: Danai Hazel Kudya — Founder & Executive Director, Africa Governance & Civic Innovation Hub (AGCIH)
AI governance may ultimately be decided by the power grid.
Discussions on artificial intelligence usually begin with algorithms, ethics, and regulation. Yet automated systems do not fail first because of law or policy. They fail because they cannot run.
As states adopt AI across administration, a foundational dependency is becoming visible: accountability depends on operational continuity, and operational continuity depends on energy stability. This shifts part of AI governance into an unexpected domain — infrastructure regulation.
Artificial intelligence is often discussed as a software question — models, data, ethics, and regulation. In practice, AI begins much earlier.
It begins with power.
Before an automated decision can be audited, trusted, or challenged, it must first run. Whether it runs reliably depends not on algorithms but on electricity stability, pricing structures, and infrastructure classification. For this reason, a quiet institutional shift is underway globally: energy regulators are becoming foundational actors in AI governance.
This shift does not originate from technology policy.
It originates from operational reality.
AI Is an Infrastructure System Before It Is a Digital System
Most digital technologies tolerate interruption. Artificial intelligence does not.
Modern AI systems operate continuously. They monitor, evaluate, predict, and sometimes act in real time. Hospitals may rely on automated triage support. Payment systems depend on uninterrupted verification. Border systems process travellers through automated checks. Administrative platforms increasingly route decisions without human delay.
Such systems do not simply pause during instability. They fail — and when they fail, accountability becomes unclear.
Electricity reliability therefore ceases to be only a service delivery condition. It becomes a condition for administrative legitimacy.
Where an automated system produces a decision — approval, denial, prioritisation, or risk classification — that decision must be reproducible and reviewable. If outages interrupt processing, institutions may not be able to explain why a result occurred or why it did not. The issue moves beyond inconvenience into the domain of due process.
In the AI era, governance depends on uptime.
The Emerging Role of Energy Regulators
This reality places energy authorities inside the operational boundary of AI governance — not by mandate, but by consequence.
1. Data Centres Become Public-Interest Infrastructure
AI requires sustained computing capacity. Governments, financial institutions, and telecommunications operators increasingly depend on data centres rather than simple office IT environments.
Energy regulation must therefore determine how such facilities are classified:
- ordinary commercial consumers
- strategic national infrastructure
- export-oriented digital services
- or a new hybrid category
2. Outages Become Accountability Events
The legal question becomes unavoidable: Can a decision be considered procedurally sound if the underlying system operated under unstable conditions?
3. Electricity Pricing Becomes Digital Industrial Policy
Artificial intelligence is energy-intensive. Tariff structures therefore shape national digital trajectories and digital sovereignty.
The Implementation Gap
Policies may define ethical principles. Sector regulators may enforce sector rules. But no single institution translates operational infrastructure conditions into governance assurance.
The function is neither advocacy nor technical deployment. It is governance architecture.
Conclusion
AI readiness is not achieved when a strategy is adopted, but when institutional responsibilities align around how automated decisions actually function in practice.
Electricity is no longer only a utility service. It becomes part of the architecture of accountability.