On August 26, Nature argued that "South Africa’s proposed registration law for scientists could become “a template for the world”. https://www.nature.com/articles/d41586-026-02568-4 The proposal deserves international attention because it confronts a question that every scientifically advanced society will eventually face: how can scientific power be made answerable without placing scientific inquiry under political permission?
Science has become too powerful to remain institutionally unaccountable. Scientific advice now shapes environmental permits, public-health decisions, agricultural systems, artificial intelligence and technologies capable of affecting entire populations and ecosystems. Yet most countries still rely on a fragmented mixture of qualifications, professional reputation, employer oversight and peer review, none of which necessarily prevents an incompetent or dishonest scientist from providing consequential advice. As an old engineer, I know that professional registration can protect the public by linking safety-critical responsibilities to demonstrated competence, ethical duties and enforceable accountability.
This regulatory vacuum has become more dangerous with the rise of organised movements hostile to scientific knowledge. Flat-earth beliefs may appear merely absurd, but vaccine denial, climate-change denial and other forms of manufactured doubt can cost lives, delay essential policies and erode the public’s capacity to distinguish evidence from propaganda. The danger becomes particularly acute when such movements recruit people with scientific credentials, because professional authority can make misinformation resemble legitimate scientific disagreement. No democratic society should criminalise ignorance, scepticism or dissent. It may, however, legitimately demand accountability when qualified professionals knowingly or recklessly misrepresent established evidence while performing activities with serious consequences for public health, safety or the environment. The ambition behind South Africa’s proposal is therefore not an attack on science. Properly designed, it could fill a regulatory vacuum that most countries have not yet acknowledged.
But a global template must draw its boundaries with exceptional care. South Africa’s proposed Bill defines scientific practice broadly enough to include unpaid advice, while failing expressly to protect research, publication, peer review and public commentary. Unregistered practice could carry imprisonment of up to twelve months, and practising scientists are not guaranteed a majority on the ministerially appointed regulatory board. These provisions risk transforming accountability for harmful conduct into political control over who may speak with scientific authority. If copied elsewhere, such a model could give governments a ready-made legal instrument for disciplining inconvenient expertise while offering the public only the illusion that registration guarantees sound judgement.
The soundest professional regimes attach registration to defined, consequential responsibilities rather than to an entire intellectual identity: in engineering, for example, responsibility for certain safety-critical designs is tied to demonstrated professional competence; in medicine, prescribing medicines is a legally restricted act. Any regime worth exporting should follow the same principle and regulate risk rather than scientific identity. Compulsory registration should apply only to precisely defined activities whose negligent performance can directly endanger people or ecosystems, for example, certifying laboratory results, signing legally consequential environmental assessments or assuming responsibility for other safety-critical scientific services. Research, publication, peer review, public-interest communication, citizen science, whistle-blowing and unpaid community advice should be explicitly protected.
The regulator should have an independently nominated scientific majority, transparent appointments and decisions, declared conflicts of interest and access to external judicial review. Sanctions should be proportional, correction and retraining for minor failures, suspension or fines for serious misconduct, and imprisonment only for intentional fraud, deliberate endangerment or persistent violation of a lawful order. Even reserved activities should be periodically reassessed against evidence of actual harm, preventing public protection from degenerating into professional closure. These safeguards would not make the law weaker. They would make it much tougher where danger is real and inapplicable where freedom is essential.
PS - This is where the argument of the manifesto, Civilizational Futures: Five Normative Imperatives for a Science-Based Civilization, meets a live legislative test. Know requires institutions capable of distinguishing reliable knowledge from incompetence and fraud, while preserving dissent and the freedom to challenge power. Care requires that scientific authority affecting people, other species and ecosystems carry enforceable responsibilities. Build requires institutions designed not only for the governments that create them, but also against future governments tempted to misuse them. A mature science-based civilization refuses to let disinformation and denial hide behind the language of dissent. When anufactured for profit or power, they become a business model whose product is preventable death. The answer is not to constrain inquiry, but to hold fully accountable anyone who wields, exploits or counterfeits scientific authority, while keeping humanity’s pursuit of truth fearless and free.
