Design Thinking for Responsible AI Product Development Training Course

Course overview

AI product teams are often asked to add generative AI, predictive models or automated decisioning to existing services before they have defined whose needs are being served, what harm could occur, or how users will challenge an outcome. The result can be weak problem framing, biased workflows, unclear accountability and expensive rework after legal, security or customer concerns emerge. This course helps professionals use design thinking to shape AI products that are useful, testable, governable and defensible from the earliest discovery stage.

Participants apply a responsible AI product-development process across empathise, define, ideate, prototype and test activities. They learn to conduct stakeholder and affected-user mapping; distinguish user needs from automation assumptions; write AI-specific problem statements; identify harms, failure modes and misuse cases; create risk-informed user journeys; design human oversight and appeal mechanisms; and translate findings into product requirements. The course connects design activities with the NIST AI Risk Management Framework and ISO/IEC 23894 so that design decisions can be evidenced in product, risk and governance reviews.

Teaching combines instructor-led demonstrations, practical workshops, case analysis and team critiques using a realistic AI product scenario. Participants work in cross-functional groups to build an evidence-based responsible AI design pack: stakeholder map, problem framing canvas, harm and misuse register, prototype workflow, testing plan, model-card inputs and implementation roadmap. They finish with a structured plan for applying the method to an active or proposed AI initiative in their organisation.

The course is suited to experienced product, design, data, risk and technology professionals who influence how AI-enabled services are scoped, built, launched or governed. It is particularly valuable where teams need a shared method for turning broad responsible AI principles into product decisions and delivery artefacts.

Course objectives

By the end of this course, participants will be able to:

• Map affected stakeholders, users and non-users using a responsible AI stakeholder map
• Frame AI product opportunities with problem statements, assumptions and measurable user outcomes
• Identify potential harms, misuse cases and failure modes through structured AI risk workshops
• Create risk-informed user journeys that expose decision points, data flows and human handoffs
• Design human oversight, contestability and escalation mechanisms for AI-assisted decisions
• Prototype responsible AI service interactions using low-fidelity screens, scripts and workflow models
• Plan qualitative and quantitative tests for usability, fairness, safety and trust signals
• Produce a responsible AI design pack aligned to NIST AI RMF and product delivery milestones

Benefits of attending

Target competencies

Who should attend

• Product Managers — who must define AI features, prioritise requirements and justify product trade-offs
• UX and Service Designers — who design user journeys, interfaces and service safeguards around AI interactions
• AI Product Owners — who coordinate model, data, engineering and business decisions through delivery
• Data Scientists and ML Engineers — who need product context for model limitations, testing and human oversight
• Risk, Compliance and Responsible AI Leads — who need practical design artefacts that support governance reviews
• Digital Transformation Leaders — who sponsor AI initiatives and need repeatable controls before scale-up

Requirements and prerequisites

Participants should have experience contributing to digital product, data, analytics, AI, risk or service-design work. Familiarity with basic AI concepts—such as training data, model outputs, generative AI, prediction, classification and human-in-the-loop review—is assumed. Participants should also understand a typical product lifecycle, including discovery, requirements, prototyping and release. No coding, statistics, model-building experience or prior formal design-thinking certification is required. A laptop with a modern browser is needed for collaborative whiteboarding and prototype exercises; access to an organisational AI system is not required.

Training methodology

The instructor introduces each design-thinking stage through responsible AI examples, then participants apply the method to a shared product case or their own approved use case. Workshops use Miro for mapping stakeholders, harms and journeys, and Figma for rapid interface and service-flow prototypes. Teams critique one another’s assumptions, controls and test plans against NIST AI RMF functions. Short case discussions examine real deployment tensions such as automation bias, opaque recommendations and contested decisions. The final session converts workshop outputs into an implementation plan with owners, evidence needs and delivery checkpoints.

Course outline

Day 1: Framing AI opportunities and affected stakeholders

• Design thinking stages for AI-enabled products
• AI product value propositions and automation assumptions
• Stakeholder mapping for users, non-users and impacted communities
• Jobs-to-be-done interviews for AI-assisted workflows
• Problem statement and how-might-we formulation
• NIST AI RMF Govern and Map functions
• AI system boundaries, data sources and decision contexts

Workshop: Participants create a stakeholder impact map and AI problem-framing canvas for a selected product scenario.

Day 2: Discovering harms, needs and risk conditions

• Responsible AI research questions and interview protocols
• User journey mapping across AI decision touchpoints
• Disparate impact and exclusion risk identification
• Automation bias and over-reliance scenarios
• Misuse case and abuse case analysis
• Data provenance, consent and purpose limitation
• ISO/IEC 23894 risk identification concepts

Workshop: Participants build a risk-informed user journey and harm register covering normal, edge and misuse scenarios.

Day 3: Ideating safeguards and accountable service flows

• Responsible AI design principles as product requirements
• Human-in-the-loop, human-on-the-loop and human-in-command patterns
• Explainability patterns for recommendations and generated content
• Contestability, appeal and correction workflow design
• Confidence thresholds and safe-failure behaviours
• Role-based accountability and escalation matrices
• Prioritising controls with impact-effort and risk-severity criteria

Workshop: Teams design a safeguarded AI service flow with oversight roles, escalation triggers and user challenge routes.

Day 4: Prototyping and testing responsible AI interactions

• Low-fidelity AI interaction prototyping in Figma
• Prompt, output and intervention touchpoint design
• Disclosure and consent microcopy for AI features
• Testing protocols for trust, comprehension and calibrated reliance
• Red-team prompts and adversarial user scenarios
• Fairness, safety and performance acceptance criteria
• Capturing model-card inputs from product discovery

Workshop: Participants prototype a critical AI interaction and run a structured peer test using safety, trust and contestability criteria.

Day 5: Embedding responsible AI into product delivery

• Translating design findings into epics, stories and acceptance criteria
• Responsible AI design pack structure and evidence traceability
• NIST AI RMF Measure and Manage functions
• Product risk reviews and go-live decision criteria
• Monitoring plans for drift, incidents and user feedback
• Cross-functional governance operating models
• Implementation roadmaps, ownership and delivery checkpoints

Workshop: Participants complete and present a responsible AI design pack and 90-day implementation roadmap for their chosen use case.

Tools & standards covered

Miro, Figma, NIST AI Risk Management Framework, ISO/IEC 23894

A typical training day

Live online deliveries follow the same structure in the East Africa Time zone, with shorter screen blocks and longer breaks.

What the fee includes

How you can take this course

Certification

Participants who complete the full five days receive the Skillset Development Certificate of Completion, stating the course title, course code, dates and delivery format — suitable for professional-development records and employer reimbursement.

Frequently asked questions