Best of 2022: 4 things to know: why data ethics is a real concern with artificial intelligence

We have been exposed to artificial intelligence (AI) in sci-fi movies and are increasingly encountering it in our everyday lives, even if we don’t realise it, both at work and at play.

And, did you know that AI is even transforming industries such as law enforcement, retail and hospitality?

AI is typically used by businesses to provide greater convenience and a seamless experience for consumers. With AI-propelled software, many business processes can be automated. For instance, managing contracts can be made easier since the AI software can help to extract information and clarify each contract's details.

In this article below, we have curated answers from an expert panel in response to participants’ questions from our webinars discussing artificial intelligence governance and the associated ethical principles.

Our expert panel includes:
• Chung Sang Hao, Deputy Director (Al Governance), Trusted Al and Data, IMDA
• Teng Chuan Hiang, Founder & Managing Partner of InterAktiv, Chairman of ERAI (Ethical and Responsible Use of Al) under APARA
• Dr David Bodoff, Senior Lecturer & Al expert, University of Haifa (Israel)
• Lyn Boxall, Director of Lyn Boxall LLC, a Singapore law firm specialising in data protection/privacy
• Moderator: Kevin Shepherdson, CEO, Straits Interactive

Look at data through an ethical lens and learn how to manage large streams of data by taking our Data Ethics and AI Governance Frameworks course here.

What are some considerations when dealing with contracts involving AI?

For contracts covering services with the AI component embedded, data privacy or protection clauses should be considered, according to Boxall. The principles of looking at the organisation’s data flows, identifying risk areas and mitigating them are the same for contracts with AI, there is no “one-size-fits-all” template. Ultimately, when organisations look at contracts, they should look at it from a risk perspective.

FAQ: What are the considerations when dealing with contracts involving Artificial Intelligence AI?

How is the integration of ethical principles done in a software architecture?

The issue of how the ethical principles are embedded inside the software development life cycle remains a challenge, said Dr Bodoff. In the engineering pipeline, it may be possible to map some of the red flags that you might think of in ethical frameworks from a privacy perspective into concrete things that can serve as trigger points. It is also possible to reverse engineer the dataset based on the outcome, according to Teng.

In the personal data field, there is a concept – data protection by design which could help technical professionals understand how to integrate principles into the development life cycle, added Shepherdson, moderator of the session.

AI Ethics FAQ: How to integrate ethical principles within a software architecture?

What is the Board of Directors’ involvement in the adoption of AI?

According to Boxall, it is the responsibility of the employee, be it a lawyer, data protection officer or engineer to communicate to the board what can be done with regards to AI and the board of directors should be the decision-maker on what actions that the organisation should take.

FAQ: What is the Board of Directors’ involvement in the adoption of Artificial Intelligence AI?

Are there any good implementation examples of the AI governance framework?

The first edition of the Model AI Governance Framework was launched in 2019 by the Infocomm Media Development Authority (IMDA) of Singapore. Upon the introduction of the Model AI Governance framework, IMDA has compiled a list of use cases detailing how organisations have implemented the framework, said Chung. The latest “Compendium of Use Cases, Volume 2” can be found here.

FAQ: What are good implementation examples of the Artificial Intelligence AI Governance Framework?

To learn more about AI governance and data ethics, register for our course here.

If you are keen to explore more data protection and privacy content, sign up for a free DPEX network membership to gain access to exclusive webinars, research and more blog content like this!

This article was originally published on 25 February 2022.