At the 18th BRICS Summit in New Delhi, South African President Cyril Ramaphosa proposed building an international oversight framework for artificial intelligence modeled on sectors already subject to rigorous regulation: aviation, pharmaceuticals, financial institutions, and above all, nuclear energy. It was an explicit acknowledgment that nuclear power, with its decades-old independent control structures, now stands as a credible reference point for anyone seeking to govern high-impact technologies.
The South African proposal calls for the creation of a BRICS mechanism for independent scientific assessment of AI systems, grounded in principles of meaningful human oversight and mandatory reporting of serious incidents. Ramaphosa also called for progressively stricter safeguards as system capabilities grow, alongside investment in the digital sovereignty of developing nations. The message is straightforward: technologies capable of endangering people’s safety cannot remain outside any regulatory framework.
The analogy with nuclear energy is no coincidence. For decades, nuclear plants have operated under the scrutiny of independent regulators, with event-reporting protocols, periodic inspections, and internationally shared technical standards. It is a system that has proven to work. Ramaphosa cited it precisely because its architecture — external oversight, transparency, and safety measures that scale with facility capabilities — is exactly what artificial intelligence currently lacks. The nuclear industry thus becomes the benchmark for a technology sector growing faster than any existing regulatory capacity can keep pace with.
Politically, the proposal fits within a broader BRICS agenda already focused on the issue. At the 2024 Kazan Summit, the bloc had adopted a Responsible AI Charter as an alternative to the European AI Act, which some member states regard as excessively restrictive for Global South economies. The New Development Bank allocated a five-billion-dollar fund for digital infrastructure in 2025. Chinese President Xi Jinping proposed the creation of an AI ecosystem within BRICS, a proposal welcomed by Pretoria. The picture that emerges is of a bloc seeking to build its own technological autonomy without abandoning a shared regulatory framework.
Ramaphosa also cited Dario Amodei, CEO of Anthropic, as an authoritative industry voice in favor of formal external oversight, independent audits, and government authority over the most advanced systems. An endorsement from an industry leader of this stature strengthens the South African position: regulation is not a brake on innovation, but the very condition that makes innovation sustainable. It is the same argument the nuclear sector has been making for years.
Should the South African proposal gain concrete traction within BRICS, the nuclear industry could find itself in an unprecedented role: not merely an energy sector, but an institutional model to be replicated for governing the technological frontier of the 21st century. It is a shift in perspective worth taking note of.
