Experts are calling for local policies and legislation that will ensure non-bias in the implementation of AI in Africa.

Hamilton Ratshefola, Country General Manager for IBM South Africa said developing rules and policies for AI systems for specific African cases can enable the development of useful AI programs, relevant to the continent.

“You’ve got to operate on principles and those principles can be fairness, inclusivity, transparency. Every company that builds AI need to have those principles,” said Ratshefola.

He added that before organisations implement an AI system, they need to know the AI ethics of the vendor.

Ratshefola said that it would be detrimental to implement an AI system that does not fully take into account the realities of the continent.

“The legal department has to look at the risks in implementing AI systems to handle customer information,” he added.

He also called on governments to implement laws that will guide the marketplace on how data on AI systems should be handled.

Currently, in Africa, few countries have implemented data laws.

The starting point for having favourable policies and ethics is to train local AI administrators and innovators.

A report by Brookings Institute early this year also called for the implementation of policies to help the development of AI in Africa.

“The region needs to formulate a comprehensive continental blueprint to guide its AI strategy by involving key Pan-African institutions, academia, and the private and public sectors in its conception,” read an excerpt from the report.

It added that: “We must harmonise regulatory policies that encourage ethically built AI systems to guarantee a more inclusive economic development for Africa. With these important steps, the next decade for Africa will be intelligent.”