As Artificial Intelligence continues to take the world by storm, adequate regulation is necessary to safeguard the interests of consumers while maintaining fair competition among service providers, the UCC Executive Director has said.
Delivering a paper on “Ethics in and Regulation of Artificial Intelligence” at the 5th Engineers Forum at Hotel Africana on 6th December 2019, Mr Godfrey Mutabazi said it’s the communication regulator’s mandate to ensure that technological innovation and consumer protection go hand in hand.
The two-day forum, which attracted the who is who in Uganda’s engineering profession, was marked by presentations and heated debates on topics ranging from industrial revolutions, engineering education in Africa, solar energy, the digital age and megacities, among others.
Going by the theme, “The Engineer in the 4th Industrial Revolution”, the forum’s principal goal was to “share and disseminate experiences, engineering knowledge, plans, readiness and solutions targeted at the 4th Industrial Revolution.”
While the world is generally fascinated by the possibilities that Artificial Intelligence promises, engineers are desperate to keep pace with the evolving technologies while regulators such as UCC are keen to protect consumers from possible harmful effects.
Mr Mutabazi said in his paper that Africa as a whole needs the right regulatory framework, “significant investment in infrastructure, and research and development to spur science, technology and innovations.”
Emerging technologies such as Artificial Intelligence (AI), Internet of Things, robotics, 3-D printing and blockchain applications, he said, present Uganda and other African countries an opportunity to leapfrog to economic transformation.
However, Mr Mutabazi pointed out, while the benefits of AI are apparent in terms of improved human productivity and efficiency, there are also fears that if not well handled, it could potentially cause harm.
“There are scares, for example, that AI could possibly result in technology becoming smarter than its creators,” he said. “How much control will we then be able to exercise on how they perform?”
Indeed, it has always been known that technology is a double-edged sword – used positively, it is a force for good, but it can also be abused and misused.
To maximise the positives while minimising the negatives, the UCC ED called for strong regulatory safeguards to eliminate or possibly mitigate risks associated with AI.
Quoting Chris Nicholson, the co-founder and CEO of Skymind, Mr Mutabazi said AI couldn’t be trusted to make perfect decisions.
“People produce algorithms, and people make mistakes,” Nicholson argues.
“Algorithms are only as good as the data that they are trained on. So if a dataset includes the historical biases of an organisation, then the predictions it makes will reflect that historical behaviour.”
At basic level, it is desirable that AI conforms to the sustainable development of society, serve humanity and champion human values that include privacy, dignity, freedom and rights.
In his paper, Mutabazi urged engineers and developers of AI “to have sufficient consideration of the potential ethical, legal, and social impacts and risks brought on by their inventions and take concrete actions to reduce and avoid them.”
On the pros and cons of regulation, he said contrary to assertions that governments should refrain from regulating technology or the Internet out of concern that it would stifle innovation, “some of the most serious challenges to open societies and open Internet today do not stem from over but under regulation of technologies.”
Nevertheless, Mutabazi acknowledged that the regulation of current and future applications of AI is not straightforward.
“As regulators, we will have to balance the desire to address legitimate social concerns, e.g. the rapid loss of potential jobs due to adoption of AI, social discrimination and security against the potential erosion of innovation and productivity,” he said, making a case for a balance between too much and too little regulation.
He also noted that as AI applications evolve over time, they may create other unique problems for regulators, giving an example of a situation where regulatory approval is mandatory before a new product or service can hit the market, yet the service is based on an algorithm that changes over time!
“How do you set service performance standards where the service continuously evolves as it is used? Such situations suggest a need for regulatory performance standards as opposed to traditional prescriptive, command-and-control regulation,” Mutabazi explained.
Uganda Communications Commission, the UCC ED emphasised, seeks to ensure that communications technology grows and thrives, but this must be done responsibly and for the benefit of consumers – not at their expense.
“Encouraging technological innovation and protecting consumers must go hand in hand and remains the core mandate of regulators,” he said.
“The digital revolution is happening swiftly, transforming traditional ways of doing every business and regulators need to be equally swift and adaptable to keep up with the pace of change, notwithstanding the complexity involved in shooting at a moving object.”