What Areas Should An AI Governance Committee Consist?

Many of you may know from my newsletter, that I am embarking on a knowledge journey on building Trustable AI.

Today I want to shine the spotlight on AI Governance. Many organizations in recently years have organized their own AI Governance Committee. The reason is because the set up of these committees sends a strong signal that the organization is focusing on building Responsible AI, which is another buzzword these days.

However, I am very caught up with this question that I have and I thought I share my thoughts on it here, as a reference for me and perhaps for someone who wants to know what areas of knowledge AI Governance Committee (AIGC) should consist of. I will be interested to hear from the readers after this, so feel free to PM me on LinkedIn!

Objective of AI Governance Committee (AIGC)

AI Governance Committee is set up to reduce the abuse and misuse of Artificial Intelligence. This is while balancing the interest of all stakeholders, allowing business owners to continue using AI for generating revenue and reduce costs through AI deployment and innovation, allowing consumers to enjoy the benefits of AI while reducing the incident of AI-assisted frauds and scams. And for governments to continue creating jobs related to AI, allowing part of the economy to be powered by AI.

With this in mind, let us get on to what are the areas of knowledge/expertise I feel should be sitting in the AIGC.

AI Expertise

This is definitely a must and its obvious they should be sitting in the committee. These are the designers, builders and engineers of AI. They are closest to the action and thus definitely warrant a representation in the AIGC. However, I feel within this, besides representation from the practitioners who are in the industry, and in addition from the academic research side.

If we are to allow innovation to flourish, the opinions from AI Practitioners and AI researchers should be represented on AIGC, its that straightforward.

Side Note: Practitioners being those that has participated in many AI Model deployment projects. The deployment experience is needed to ensure the policies produced allows the smooth deployment process, from data collection, pipeline building, model training, model deployment, tech stack design and feasibility, to designing the automated decision- making process.

Economics & Policy Making

The product from any AIGC will either be a set of recommendations (associations and businesses) or policies (government). Policy making is a very delicate process there is akin to a brain cancer surgery. Precision and accuracy is very important, so as to ensure the bad cancer cells are removed while the brain is intact, in a perfect scenario.

Having someone who has policymaking background will understand how to balance the interests of many while allowing innovation to flourish. Policymaking background helps in having a good understanding on how the stakeholders will be impacted when policies are released.

For most policymaking procedures, having some understanding of Economics will help in ensuring that there is positive feedback from the policymaking into business and the macro-economy.

Jurisprudence and Lawyers

"Jurisprudence is the philosophy and theory of law. It is concerned primarily with both what law is and what it ought to be. Jurisprudence provides a critical framework for understanding the nature, purpose, and functioning of law. It encompasses various schools of thought, each offering unique insights into the relationship between law, society, and morality."

~ From Wikipedia & ChatGPT

From this definition, I am very sure my readers will know why they should be included in the AIGC. Policies consist of laws, guidelines and assistance in interpretation of keywords. The set up of the laws and guidelines will need jurisprudence background to understand how the laws set up will impact the society and more importantly the morals of individuals, in turn leading to the ethics of the society.

Setting up these laws and guidelines can be a make a break on whether we can ensure innovation to flourish while reducing the intended and unintended abuse and misuse of Artificial Intelligence.

After jurisprudence, comes the lawyers who are going to work within the policy framework. AIGC will need their professional knowledge on the writing, and the words used in writing the policies. Lawyers practicing intellectual property law, criminal law, contract law will be needed in the short term practicing labor & employment law, and environment law will be needed.

With the assistance from practicing lawyers, the wordings on the policies can be tightened to reduce its ambiguity and in turn allow innovation to flourish within the well-interpreted governance framework.

Regulators & Auditors

A strong governance framework without regulators and auditors to enforce it will just be weak in execution. But audit checks must be structured in a way that minimally it can be checked qualitatively, and best to be checked quantitatively. To structure it properly so that it is enforceable, auditors warrant a representation on the committee. This allows the governance process to be closed off properly, from structuring policies, guidelines and laws, enforcing it, to clear-cut evidence on abuse and misuse of Artificial Intelligence.

Business Owners

Of course not forgetting the centerpiece of the AIGC and that is the business owners. They definitely need to be represented in the AIGC, because they will be the primary user of Artificial Intelligence. Any economy will want the business owner to be able to use Artificial Intelligence to generate massive value for the economy, create AI-related jobs, pay part of the massive value to government as taxes. All these benefits while increasing the likelihood of abuse and misuse being caught, together with reducing the incentives to abuse and misuse AI.

Consumers!

And of course how can I forget the consumers that will be impacted by the decision made by Artificial Intelligence. They definitely need to be represented, as for how to choose the representation...I am kind of stumped right now. For sure the minorities need to be represented but we do not want to bloat up the AIGC. This is because there is a tradeoff between the size of the committee vs the effectiveness and efficiency of the AIGC. Maybe you can share your thoughts on this part with me.

I will pause the article here for a while and update it in a later post perhaps.

Concluding Thoughts

In my opinion these are the main players on AIGC. Depending on where the AIGC is residing, whether within a business or non-profit organization, or at the country level, there will be a need for further "customization" depending on which area of AI the organization or government is going with.

From here, readers can see that setting up a good AI Governance Committee is not so straightforward. There is a need to see if there are people with the relevant background, and if there is decent size pool of these talents, who to select and whether the invitation will be accepted. And not forgetting how the chairman and vice-chairman should be selected followed by how the meeting should be conducted and how final decisions should be made.

There is definitely a lot of work to be done but it needs to be done to at least ensure that Artificial Intelligence, perhaps the final tool that humankind will be building, does not get abuse, misuse and bring harm to society and humanity.

Thanks for reading till here! I know it is a long post, as I eye the word count at around 1,200 words!

What are your thoughts? Consider sending them to me on LinkedIn!

Support my work please! Share this article with your network or buy me some "books" to support! Or subscribe to my newsletter below for short form of my thoughts. :)

Note: Image was created by Bing Image Generator.