By Moroasereme Ntsoane — CEO, Summit AI
A New Threshold of Governance
With the launch of King V Code of Corporate Governance by the IoDSA on 31 October 2025 (effective 1 January 2026), one can truly say the corporate governance landscape in South Africa has entered a new chapter. The Code explicitly recognises the transformative role of Artificial Intelligence (AI) and other disruptive digital technologies in shaping corporate life.
Far from being a light touch, the current update of the Code has significant implications for corporate governance. King V draws a clear line between the King IV world that boards once governed, and the one they must now learn to steward. In this new world, intelligent machines inform business models and strategy, big data rivals capital in its strategic importance, and decisions made by algorithms come with deep human and ethical consequences.
Fiduciary duty in the age of intelligent machines is no longer just about financial prudence. It now includes ethical, trustworthy, and human-centred oversight of intelligent technologies. The implications of all this couldn’t be more profound for any board.
From Compliance to Stewardship
King V is a strong reaffirmation of the principle that the purpose of corporate governance is not mere baseline compliance, but a commitment to create sustainable value within an organisation’s economic, social, and environmental milieu. This framing places technology governance, and specifically AI, at the core of strategy and risk, rather than allowing it to be seen as a voluntary good practice or complex operational challenge.
The Code demands that boards exercise foresight by transforming from being passive observers of technology deployment to becoming courageous stewards of algorithmically-powered outcomes. This is the moment for directors to demonstrate the ethical leadership that King V requires.
The Ethical Imperative
The previous King IV Code also addressed technology and information governance. However, King V significantly refines and elevates this principle by explicitly addressing AI systems and their ethical dimension. Principle 10 of the Code – dealing with data, information and technology – stipulates that:
“The governing body governs, data, information, and technology in a way that enables the organisation to sustain and optimise its strategy and objectives.”
Although concise, this principle is indicative of a major shift in corporate governance practice within boards. The board’s responsibility for due care now incorporates the obligation to understand, oversee, and be accountable for the organisation’s digital and algorithmic dependencies.
Specifically, boards of companies must ensure that:
oAI systems and data practices within the organisation uphold universal human rights and fairness.
oOversight and governance frameworks include mechanisms for human review and override of automated decisions, i.e. human-in-the-loop.
oAccountability for algorithmic outcomes is fairly allocated, traceable, and defensible.
oAssurance over data ethics, privacy, and security is fully compliant with all legal prescripts, and also included within governance reporting.
In short, directors are expected to treat the digital infrastructure of their organisations with the same rigour with which they handle financial statements and audit integrity.
The Convergence of Law, Ethics and Technology
Globally, the regulatory landscape for AI technologies is gradually hardening, driven by frameworks like the EU AI Act, which imposes clear risk-based legal obligations on developers and deployers of AI systems. Locally, however, the formal regulatory environment remains in its nascent stages, and is currently characterized by the National Policy Framework for AI, which is still in draft, and with comprehensive and binding legislation expected to follow. This criEcal regulatory gap means that the burden of responsible AI deployment, risk mitigation, and ethical assurance falls squarely on the shoulders of the board through their fiduciary and governance duties.
By embedding global values like fairness, accountability, and explainability within Principle 10, King V strengthens South Africa’s position as a governance pioneer in the Global South. It serves as the primary enforceable ethical and control mechanism in the absence of a dedicated AI law, recognizing, as it does, that the legitimacy of business in the digital era depends on the trustworthiness of its intelligence systems.
This alignment also means that boards cannot rely on ignorance or outsourcing of their fiduciary commitments anymore. Fiduciary awareness now extends to understanding how algorithms learn, what data they consume in the process, and whose interests they may inadvertently harm. Therefore, directors must acquire the literacy to ask the right questions about these models, and the humility to know when to request specialised guidance to augment their endeavours.
Fiduciary Duty in the Age of Automation
Traditionally, fiduciary duty has meant acting in the best interests of the organisation with care, skill, and diligence. King V reframes and extends this duty to accommodate a world where data-driven systems shape every decision. The principle of due care now includes the obligation to understand, at least in essence, the organisation’s digital and algorithmic dependencies.
Boards must therefore integrate AI governance and oversight across all verEcals, into:
oRisk governance: identifying AI-related risks such as bias, misinformation, or systemic discrimination.
oEthics: ensuring that AI behaviour aligns with organisational values, social expectations, and norms.
oCompliance: confirming adherence to POPIA, the Cybercrimes Act, and forthcoming AI regulations.
oAssurance: obtaining independent verification that AI systems meet ethical and legal standards (Trustworthy AI).
This holistic approach transforms fiduciary duty from a backward-looking audit function within the organisation into a forward-looking act of stewardship.
Ubuntu as the Guiding Philosophy
At the heart of King V’s treatment of AI models is a simple moral insight that dictates that technology must remain accountable to people whose interest it purports to serve. Therefore, boards are urged to ensure that decisions generated by these algorithms remain ultimately answerable to human judgment (human-centred). Therefore, explainability, transparency, and fairness are not merely technical requirements, but they are expressions of Ubuntu as the human-centred philosophy, which is incorporated within the Code.
In this way, King V blends South African ethical heritage with cutting-edge governance practice, and reminds us that leadership in the digital era is not about controlling machines, but guiding humanity’s relationship with them, and making sure that they serve humanity instead of the opposite.
Looking Ahead
As King V becomes effective in the beginning of the new year, for most entities, the window to engage these governance foundations is narrow. Yet this is not a deadline to fear; it is an invitation to lead. Boards that begin the work now will not only demonstrate compliance, but will set the ethical standard for a continent that is proactively entering its fully digital decade. Those who start this journey today will define what ethical and trustworthy corporate leadership means in the age of intelligent machines.
This shift requires more than aspiration; it requires a strategic governance framework to translate Principle 10’s mandate into an auditable reality. This is where expertise becomes essential.
As the line between human and artificial intelligence continues to blur, the quality of governance will determine whether technology amplifies human values or erodes public
trust. Failure to define these controls now could expose the organisation and its directors to significant legal risk and also threaten long-run stakeholder trust.