Unlocking AI Value Responsibly: Why boards must act before year-end
View(s):Feature
By Ruwani Dharmawardana
Artificial Intelligence (AI) has transcended the realm of futuristic speculation; it is now a strategic imperative demanding immediate board-level attention. To relegate AI to next year’s agenda is to overlook its profound capacity to reshape operational paradigms, accelerate innovation, and redefine competitive advantage. Organisations can start with enterprise AI gradually, focusing on incremental use cases that drive efficiency and innovation without requiring major budgets or disruptive system changes. Yet, this promise is interwoven with nuances and intricacies that boards must navigate with discernment. The mandate is not to stifle adoption but to enable it responsibly, embedding AI within a governance architecture that harmonises ambition with accountability.
AI is revolutionising how leaders and teams manage complexity, but its power must be guided responsibly. Board members can receive concise summaries of lengthy board papers before meetings, saving hours of preparation, while CEOs leverage AI to draft strategic briefs or scenario analyses in minutes. CFOs gain rapid financial insights, company secretaries can turn meeting transcripts into accurate minutes, and auditors accelerate risk reviews with AI-assisted checks. Sales teams prioritise leads, and call centre managers analyse customer sentiment instantly. These capabilities are real and accessible today, but they demand human-in-the-loop oversight, especially as organisations experiment with automated decision-making in areas like credit approvals, recruitment screening, and compliance alerts. Boards must ensure these systems remain transparent, explainable, and subject to human override to prevent bias, errors, and regulatory breaches.
Achieving this balance calls for a truly cross-functional approach, aligned with the organisation’s defined and regularly revisited risk appetite and tolerance levels. Boards set the strategic vision and provide oversight, while ICT and data teams build secure, scalable systems. Compliance and legal functions ensure regulations and contracts are met, risk managers anticipate and mitigate emerging threats within acceptable boundaries, and both internal and external audit update their programmes to include algorithm audits, data governance checks, bias and fairness testing, explainability reviews, cybersecurity controls, and assurance that human-in-the-loop oversight is maintained. Agile practices must also be embedded to adapt quickly to regulatory changes and emerging risks. Together, these roles create a unified governance framework that turns AI into a responsibly managed strategic asset driving innovation while staying within agreed risk parameters.
AI exists across a spectrum of capability. At the foundational level is ‘Narrow AI’, systems built for a single purpose, such as Siri or Alexa handling voice commands, Netflix’s recommendation engine tailoring content, or spam filters protecting email inboxes.
A step further is ‘Advanced Narrow AI’, which remains domain-specific but demonstrates remarkable sophistication. Examples include ChatGPT, Microsoft Copilot, Google Gemini, Anthropic Claude, Perplexity AI, China’s ERNIE Bot, and DeepSeek. These systems can generate text, summarize documents, translate languages, assist with coding; enable computer vision for autonomous vehicles; and power scientific tools like AlphaFold for protein structure prediction.
Beyond these practical applications lies ‘General AI (AGI)’, still hypothetical, which would mirror human cognitive flexibility, an AI capable of designing buildings, managing finances, and learning new skills without retraining. Some multi-modal AI systems (combining text, image, audio, and video) and agent-based architectures are moving toward AGI-like behaviour, but they remain domain-constrained.
The ultimate frontier is ‘Super AI’ (ASI, with even post-ASI possibilities), a theoretical stage where machines surpass human intelligence entirely, a theoretical stage where machines surpass human intelligence entirely, innovating and governing at scales beyond human comprehension. While AGI and ASI remain aspirational, Narrow and Advanced Narrow AI are already transforming corporate operations, regulatory compliance, and government service delivery, bringing both unprecedented opportunities and governance challenges that demand immediate board-level attention.
The EU AI Act, along with emerging national laws, OECD principles, and country-level AI strategies, is shaping a global regulatory landscape for responsible AI adoption. Together, these frameworks create a compliance environment that demands proactive data governance, AI inventorying, risk classification, vendor oversight, explainability, and robust accountability processes well before fully automated systems are deployed.
(The writer is an experienced Sri Lankan lawyer and corporate expert based in New Zealand. She is hosting a Zoom workshop in December on responsible AI adoption. The session will cover technical, legal, governance, and risk aspects, offering practical insights for boards and senior leaders. Those interested in joining can reach out via email: rulegalthoughts@gmail.com).
Hitad.lk has you covered with quality used or brand new cars for sale that are budget friendly yet reliable! Now is the time to sell your old ride for something more attractive to today's modern automotive market demands. Browse through our selection of affordable options now on Hitad.lk before deciding on what will work best for you!