Logo Are You A Robot


Bridging the AI Gap

Navrina Singh

In our interview with Navrina Singh, we dive into Credo AI and what good AI governance should look like. We also think about the role humans play when developing tech and ensuring there is always a human oversight in these processes.

Navrina is the founder and CEO of Credo AI, Mozilla Board Member and Young Global Leader at World Economic Forum. You can follow her on Twitter or LinkedIn

AI is continuously advancing at a fast pace. When it comes through the market, AI starts to influence different areas of the market. In that development speed, we have not been putting guardrails in place for when things go wrong. Navrina argued in the pace and excitement of AI development, an ‘oversight deficit’ has been created due to the lack of continuous conversations between data scientists and policy and compliance business departments.

Navrina also highlighted the importance of ‘model scorecards’ on AI applications. As consumers, we are never buying anything without understanding the label and context. Like nutrition labels at a supermarket, AI model scorecards allow users to identify what type of data was used to train the model and techniques. These model cards can also be used for different stakeholders; for example, one specific group might be interested in the metrics, whereas another is solely on the audit processes. This allows for complete transparency across the organisation and an understanding of the structures of the enterprise.

  1. AI governance overall is still in the early stages. Within the enterprise, many companies are asking questions such as:
  2. Who is responsible?
  3. Should we wait until regulation?
  4. How can we make AI governance a seamless process?
  5. How do we ensure persona metamorphosis is happening?

Navrina explained that persona metamorphosis is essential to create a more holistic governance structure. This is where personas from specific groups, for example, risk and data science departments, bring in their expertise and learn about AI and compliance to think about what new governance structures are needed to implement it into the company culture.

What are your thoughts? Join our Slack channel and join the conversation!