Startup accelerator programme

Scaling AI globally is no longer limited by technology, it is limited by ethics and trust.

For founders, the real barrier to international growth is not building AI, but proving it can be governed, trusted, and deployed responsibly across jurisdictions.

This session reframes governance from a compliance burden into a strategic growth engine. Drawing on globally recognised frameworks, including the EU AI Act, ISO/IEC 42001, and NIST AI Risk Management Framework, it translates regulatory complexity into practical, founder-ready actions.

Through real-world examples and application-level insights, founders will learn how to move from fragmented AI experimentation to structured, scalable governance.

The session challenges the assumption that governance slows startups down, and shows how it accelerates enterprise adoption, investor confidence, and cross-border expansion.

Learning objectives:

By the end of this session, founders will be able to:

1. Analyse how ethics, trust, risk, and regulation impact AI scalability across global markets.

2. Evaluate initially how the EU AI Act, ISO/IEC 42001, and NIST AI Risk Management Framework apply to their business models.

3. Design a practical, application-level AI governance approach to support international expansion.

4. Apply governance, trust, and compliance principles to strengthen investor readiness and enterprise sales.

5. Assess their AI governance maturity and prioritise actions for global growth.

AI projects promise to transform education, health, and livelihoods... but how do you know if yours actually works? Most AI teams know how to run technical evaluations: testing accuracy, safety, and performance. Far fewer know how to measure what actually changes in the world: the behaviors, outcomes, and lives their project is meant to improve. This session closes that gap.

We'll explore what the field has built so far, drawing from the development sector, the UN ecosystem, and the corporate world, and comparing approaches across two essential questions: does your AI work, and does it matter? We look at what each approach measures, when it applies, and what it misses, so you can choose, combine, or build what fits your context.

Whether you're building a health chatbot or a predictive tool for climate risk, you'll leave with a map of where your project stands and what to measure first. Measuring and reporting isn't just about accountability. It's about how you improve, iterate, and gather evidence that drives action. The field is still being built. Your project can contribute to it.

Learning objectives:

By the end of the session, participants will be able to:

1. Distinguish between technical AI evaluation and social impact measurement, explaining why each is necessary, what each misses, and why most AI for Good projects rely on only one.

2. Compare existing approaches to AI project evaluation across the development sector, UN standards, and corporate frameworks, identifying what each captures, when it applies, and what it leaves out.

3. Evaluate the fit between different measurement approaches and their own project's stage, context, and available data, and identify a concrete first step.

4. Apply measurement findings into a reporting narrative that drives action.

Most early-stage AI ventures fail not because the idea was wrong, but because the wrong decisions were made too late, the wrong ones too quickly, and the founder ran out of capacity before the mission scaled.

This session is a practical operating guide for founders building AI for social good. It addresses three questions that determine survival: which decisions are reversible and which are not, how to build personal and organisational resilience without slowing down, and when to scale, when to deepen, and when to deliberately stay small.

The session shares concrete patterns from founders who endured and from those who did not. The focus is on what founders rarely admit out loud: the cost of the wrong decision, the price of permanent urgency, and the discipline required to invest in your own capability while everyone else is asking you to ship faster.

Founders will leave with a usable framework for classifying decisions, a personal resilience practice, and a clearer sense of which scale problem they are actually trying to solve.

Learning objectives:

By the end of this session, founders will be able to:

1. Distinguish between reversible (two-way door) and irreversible (one-way door) decisions in their venture, and apply the appropriate decision speed to each.

2. Evaluate which signals in their market and technology landscape are durable versus temporary, and prioritise the assumptions most worth testing.

3. Design a personal and team resilience practice that protects creative judgment under sustained uncertainty.

4. Determine when to scale operations, when to deepen impact, and when to remain deliberately small to protect mission integrity.

5. Construct a structured plan for ongoing capability investment in themselves and their team, calibrated to the pace at which AI is now evolving.