Empowering teams with clarity, consistency, and lasting impact across sectors.

Ethical AI requires testing, learning, and refining

Jeffrey K. Liker’s interpretation of the Toyota Production System (TPS), or as he calls it, the Thinking Production System, emphasizes process as a core driver of quality and reliability. Toyota’s teams follow a structured cycle of planning, experimenting, checking results, and improving. This is famously captured in Plan, Do, Check, Act.

Adapting the methodology to the organizational use of AI, it is safe to say that ethical AI use demands the same mindset. AI systems evolve as data, context, risks, and user behaviour evolve. A one time ‘compliance check’ is not enough. Leaders must treat AI governance as an iterative process, not a static policy.

This is where Process connects directly to Philosophy (Toyota’s first pillar). A clear purpose establishes why an organization uses AI; a disciplined process determines how well it is used. Purpose without process becomes aspiration with no structure. Process without purpose becomes efficiency without ethics. When combined, they ensure AI systems not only serve a meaningful mission but are continuously assessed, challenged, and improved to stay aligned with that mission over time.

A strong AI processes include:

  • Planning with clear purpose, safeguards, and expected outcomes;
  • Conducting responsible development, testing, and deployment;
  • Checking for accuracy, fairness, privacy, security, and unintended impacts;
  • Acting on those findings to refine, update, or redesign the system.

This cycle repeats, not occasionally, but continuously and Toyota’s pillars remind us that responsible innovation is not accidental, it is intentional, structured, and learned over time. Philosophy gives us the why. Process gives us the how. Together, they create the foundation for safe, meaningful, and trustworthy AI use.

Having said that, philosophy and process only work when they are translated into real organizational practices, the very purpose of the AI-RESPECT™ (patent pending) Compliance Framework. AI-RESPECT helps organizations define purpose, build repeatable processes, and embed continuous improvement into everyday workflows. It transforms responsible AI from a one time compliance task into an operational discipline.

AI is dynamic, and so governance must be, too. By adopting an iterative approach grounded in clear purpose and disciplined process, organizations can ensure that the AI systems they use remain aligned, accountable, and continually improving just as the Toyota Production System intended.

-CT