As organisations across South Africa and the broader continent race to adopt AI, from customer service chatbots to fraud detection models to generative tools embedded in daily operations, few have paused to ask who is governing these systems, what data they touch, and what happens when they fail. This keynote will argue that ungoverned AI adoption is quietly becoming the next major cybersecurity crisis: a threat vector hiding in plain sight, expanding faster than most institutions' policies, oversight structures, or regulatory frameworks can keep pace with. Drawing on BCF's work advising government, healthcare, and financial services clients across U.S. and African markets, Dr Robinson will make the case that AI governance is not a compliance afterthought, it is now core infrastructure security.

For a South African audience navigating POPIA compliance, sector-specific regulation, and an accelerating AI adoption curve, this keynote connects global lessons in AI risk to local realities, equipping leaders to build governance structures now, before incidents force their hand.

• Shadow AI is the new shadow IT — Employees and departments are adopting AI tools faster than security and legal teams can vet them, creating unmonitored data flows and unmanaged risk across the enterprise.
• Data governance is AI governance — For South African organisations, POPIA compliance and AI oversight are inseparable; feeding sensitive or personal data into ungoverned models can quietly turn a productivity tool into a regulatory and security liability.
• AI systems are attack surfaces, not just tools — Model manipulation, data poisoning, and prompt injection are real, emerging threats; treating AI as "just software" ignores risks that traditional cybersecurity frameworks weren't built to catch.
• Governance Is a competitive advantage, not a constraint — Institutions that build clear AI policies, oversight committees, and accountability structures now will move faster and more safely than competitors scrambling to retrofit governance after a breach or regulatory action.

As organisations increasingly adopt AI, intelligent automation and autonomous business processes, boards are expected to provide effective oversight of technologies that can influence decisions, behaviours and risk exposure at unprecedented scale. While many organisations have established AI principles and policies, directors often lack practical mechanisms to determine whether technology is being used ethically, responsibly and in a manner that maintains stakeholder trust.

This presentation shares a South African case study on the development of an Information Services Ethical Behaviour Assurance Index (IS-EBAI), a board-level governance model designed to provide measurable assurance over the ethical use of technology, data and AI. The framework consolidates ten ethical governance dimensions into a single board reporting structure, including privacy, integrity, accountability, transparency, fairness, responsible AI and third-party stewardship.

The presentation demonstrates how complex technology and AI risks can be translated into meaningful board assurance indicators, supported by evidence-based reporting, governance thresholds and escalation mechanisms. Delegates will gain practical insight into strengthening board oversight of AI-enabled decisions, balancing innovation with trust, and preparing for the evolving future of technology governance in South Africa.

Key Delegate Takeaways

• A practical approach to transforming AI and technology ethics principles into measurable board assurance.
• A reusable governance framework that enables directors to oversee intelligent enterprises with greater confidence.
• Practical lessons on balancing innovation, competitiveness and responsible technology governance while maintaining organisational trust.

Cybersecurity and operational risks are no longer purely technical concerns, they can directly impact revenue, productivity, compliance, customer trust and business continuity. This session explores how organisations can translate technical risk into measurable business and financial impact, giving executives and boards a clearer view of where exposure exists and where investment can reduce risk. Using ManageEngine’s security, IT operations, endpoint, identity and observability capabilities, delegates will see how unified visibility and actionable data can support better risk prioritisation, resource allocation, and informed board-level decisions.

• Translate technical risk into business impact; Connect security incidents, vulnerabilities, downtime, and operational disruptions to potential financial and business consequences.
• Build a complete view of organisational risk; Combine insights from SIEM, identity, endpoint, vulnerability, network, and application monitoring to identify risks across the IT environment.
• Prioritise investments based on risk exposure; Use real-time visibility and analytics to identify critical assets, vulnerabilities and operational dependencies that require attention first.
• Give the board actionable risk intelligence; Move beyond technical alerts and dashboards by providing meaningful metrics around exposure, compliance, resilience, downtime, and potential business impact.

The session will explore how organisations can scale AI without losing accountability, transparency, or human judgment. Drawing on practical governance experience and global developments, it will challenge leaders to rethink how trust is built in an increasingly autonomous world.

• Why traditional governance approaches are struggling to keep pace with agentic AI
• The difference between having humans "in the loop" and ensuring meaningful human oversight
• Balancing innovation, automation, and accountability
• Practical approaches for boards, executives, and governance leaders to manage AI-enabled decision-making
• Building trust as a strategic capability rather than treating governance as a compliance exercise

Employees are already pasting sensitive data into AI tools, and agents are starting to act on their behalf, often with no one having approved either. Most organisations can’t say which agents are running or who is accountable for them. This session treats AI governance as an accountability question, not a tools problem. It shows why a named human must answer for every agent, how a cross-functional council can say “yes, if” instead of just yes or no, and why seeing where data actually moves is the first step to governing it

South Africa’s digital ambitions depend on the capacity of its institutions and leaders to govern technology with sound judgement, clear accountability and public trust. As AI and increasingly autonomous systems reshape organisations, governance capability must keep pace with technological adoption.

This closing keynote will connect the conference’s central themes to the practical capabilities boards and executives need: asking informed questions, assigning decision rights, obtaining credible assurance and building cultures where people challenge assumptions and escalate concerns. Grounded in South Africa’s realities, it will explore how organisations can support responsible innovation while strengthening resilience, inclusion and long-term value. Delegates will leave with a focused agenda for turning the day’s insights into action.

• Identify the capabilities that matter: Assess gaps in board and executive digital literacy, judgement, ethical leadership and oversight.
• Make accountability explicit: Clarify who approves, monitors and intervenes in AI and digital initiatives, including those involving external providers.
• Turn trust into evidence: Ask sharper questions and seek meaningful assurance on data use, cyber resilience, AI outcomes and stakeholder impacts.
• Translate insight into action: Develop a focused governance improvement agenda with clear priorities, owners and evidence of progress.