Establish data ethics in Artificial Intelligence

Ethical aspects of artificial intelligence (AI) remain rife and raise questions in many quarters. Yes, AI comes with attendant risks. But also presents immense opportunities for businesses and individuals alike. There must be a balance/marriage between ethics and data usage. A PwC research claims AI opens virtually limitless potential to benefit the whole of society and estimates that business investment could translate to an infusion of $15.7 trillion into the global economy by 2030.

Meaning that in the future, companies would rely heavily on analytics and drawing reports to diagnose problems and come up with the best possible solutions. Imagine a housing estate that wants to improve security and proceeds to place CCTVs, gather enough data about incoming and outgoing, and other general activities. However, there are certain implications to doing this, first would be the warehouse where this data was stored — how was it built and what were the IT architectural considerations put in place? What are the policies, who will implement access controls, what are the acceptable use policies, what are the retention and deletion protocols?

Organisations mostly focus on the use of data to improve their business processes, build intelligent products that can, for example, predict customer behaviours and add to the overall revenue base of the company. But there is always the question of how ethical is the procedure? What steps can be dangerous? How can the end-2-end circle of data within this ramification be managed?

A recent Responsible AI diagnostic, surveyed around 250 senior business executives from May to June 2019, which found that the level of understanding and application of responsible and ethical AI practices among respondents significantly varied across organisations, and in most cases was immature.

The findings also highlighted critical challenges around the skills needed to ensure there are responsible ways for AI practices and the corresponding response that about 25 percent have not considered AI as part of their corporate strategy, only 38 percent well think it is aligned with their organisational values and only 25 percent prioritise a consideration of the ethical implications of AI solution before investing in it. As we can see the number is low when it comes to ethical consideration for most companies.

The reason being that most companies have not been able to think about the regulatory implications and the excessive intrusiveness this presents when it comes to the use of data and manifestation in artificial intelligence. The future is going to run on data and the companies that want to succeed understand that the collection of data sets will play a pivotal role in the making or breaking of the company.

With that in mind, I propose a simple 7-step approach when it comes to ensuring that ethics and data usage are well merged to produce the right results. The approach includes 1. Strategy 2. Set up ethics committee 3. Planning 4. Development 5. Testing 6. Deployment 7. Operations and monitoring. This is to highlight for stakeholders to consider and use in accordance to the structure of their business operations or projects.