Many organisational AI conversations begin with a deployment question: which tool should we adopt, and where can it save time? De Kai, an AI speaker, invites a more demanding question: what are we teaching the systems that are increasingly participating in communication, decisions and work?
His central idea is that AI is not simply a passive tool. Through the behaviours people reward, the data and feedback they provide, and the incentives organisations create, humans are in effect “raising” AI. That framing sits at the heart of Raising AI, which is also available as an audiobook. For leaders, the idea shifts responsible AI from a late-stage compliance exercise to an ongoing question of organisational culture and stewardship.
What does it mean to “raise” AI?
AI parenting is a useful way to describe a continuing responsibility rather than a one-off technical task. Organisations do not merely select a model, set it live and step away. They shape how AI is used through policies, performance measures, escalation routes, acceptable-use norms and the examples employees see from senior leaders.
This matters especially as organisations experiment with more agentic AI: systems designed to take actions, pursue tasks or operate with a degree of autonomy within defined boundaries. The more a system can influence workflows or customer interactions, the less sufficient it is to ask only whether it functions technically. Leaders also need to ask what behaviours it may reinforce, whose interests its incentives serve, and when people must intervene.
De Kai’s argument does not suggest that AI independently absorbs every feature of an organisation. Rather, it directs attention to the human conditions around a system: the goals chosen for it, the feedback surrounding it, and the standards used to judge its outputs. Those conditions are leadership decisions.
Responsible AI leadership starts with organisational behaviour
A useful implication for boards and executive teams is that AI governance cannot sit solely with technology or legal functions. Human resources, learning and development, risk, operations and frontline teams all influence the environment in which AI is adopted.
Leaders can turn the concept into practical governance questions:
- What behaviour are employees being rewarded for when they use AI: speed alone, or accuracy, judgement and appropriate challenge?
- Which decisions may be supported by AI, and which require accountable human review?
- How can staff report harmful, inaccurate or inappropriate outputs without fear of slowing delivery?
- Do teams understand the limits of the systems they use, including when not to rely on them?
- Are customer, employee and societal impacts considered alongside efficiency and productivity goals?
These questions make AI culture visible. A company that measures only volume and turnaround time may encourage very different AI use from one that also values verification, transparency and escalation. Neither outcome is determined by software alone.
From AI policy to everyday practice
Policies remain important, but they can be too abstract to guide daily decisions. The “raising” metaphor points to repetition: people learn from what leaders model, what systems make easy, what managers tolerate and what teams discuss after something goes wrong.
For example, an organisation introducing AI assistance for customer communications might define when employees must check outputs, how sensitive information should be handled, and what language is unsuitable for automated drafting. It might also review real examples from use, update guidance and make responsibility clear when an AI-supported interaction causes concern. The point is not to eliminate all risk; it is to create habits of informed human oversight.
For HR and L&D leaders, this creates a learning agenda that goes beyond prompt-writing. Employees may need practice in assessing reliability, identifying bias or inappropriate assumptions, protecting confidential information, and recognising when a high-stakes matter should be escalated to a person with authority.
Why the conversation is timely for Asia-Pacific audiences
De Kai brought this perspective to IDECS26 in Kuching, where he delivered the opening keynote at the Centre of Excellence track, “From the Lab to a Billion Users: What Translation Taught Me about Responsible AI Commercialisation.” Drawing on his experience in machine translation, he explored the practical challenges of taking AI innovation from research to global-scale use. In that context, his human-centred perspective offered an important message for business leaders: technical capability creates value only when organisations also design the right guardrails, incentives and human oversight.
For conference organisers, this makes De Kai relevant to programmes on AI governance, digital transformation, executive responsibility, workforce readiness and technology ethics. Rather than treating responsible AI as a specialist topic detached from strategy, a session can help leaders connect it to the choices they make about culture, incentives and accountability.
Questions worth taking into the next AI strategy meeting
De Kai’s Raising AI framework gives leadership teams a clear starting point. Before asking what an AI system can do, ask what the organisation is teaching it to optimise for. Before expanding autonomy, decide where human judgement remains essential. And before presenting AI as a neutral productivity tool, examine the behaviours and values that its use may normalise across the business.
That is the practical force of AI parenting: responsible AI is not only about controlling technology. It is also about the kind of organisation people are building around it.
To enquire about De Kai for a conference, leadership meeting or corporate event, contact Speakers Connect at info@speakersconnect.com.

