What is responsible AI?

What is responsible AI? A quick glance at this core topic from a corporate governance perspective

There’s no denying that artificial intelligence (AI) is transforming the corporate landscape at an unprecedented pace. It can automate, it can process huge swathes of data, and it can function as the assistant that many in the governance space never had before. 

Alongside that potential, you’ve probably heard an extensive amount about the responsibilities that come with using AI. Critics point out that you need to be able to catch bias, poor data, wrong information, and other complex drawbacks. 

While AI might seem like a distant issue at times, especially to people in the boardroom, they must play an active role in the ethical and operational issues around AI, ensuring it serves business and society according to modern, strict standards. 

Responsible AI is a set of principles designed to deploy AI systems in a way that is ethical, fair, transparent, and accountable. In other words, it places the notion of “common good” alongside business success in the rollout of any AI-based solution. 

Practically, responsible AI frameworks will likely include measures to mitigate biases, protect data privacy, and ensure transparency in decision-making. 

While responsible AI frameworks will differ across countries and industries, sometimes even across individual companies, the underlying ethical compass remains largely the same.

It’s essential because stakeholders care about it. 

AI has been around for years in some form, but attention skyrocketed with the release of Gen-AI models like ChatGPT in the early 2020s. Critics immediately began to ask how those in power would be able to control the vast new potential of AI models. Quickly, these concerns fed through to stakeholders like investors, consumers and employees, who wanted clarity on how each AI system would work, how it would impact people, and how any flaws would be contained. 

The desire for this clarity hasn’t faded since.

First and foremost, boards should never consider themselves apart from AI governance. It’s not an integral part of boardroom responsibilities. Think of it this way: directors have a fiduciary duty to oversee risk and strategy; this duty extends directly to AI. 

Here’s what to focus on:

• Oversight and governance: Boards should ensure a clear AI governance framework is in place. This includes setting a high-level strategy for AI adoption and ensuring there are policies to govern its development and use.
• Risk management: Directors need to understand the risks of AI, including algorithmic bias, data security breaches, and potential regulatory non-compliance. They should be asking questions as much as possible, especially now, as AI remains new and evolving at a rapid pace. 
• Culture and ethics: Board members are responsible for fostering a company culture that prioritises ethical considerations in AI development. This starts from the top and must be embedded throughout the organisation.
• Accountability: Ultimately, the board is accountable for the organisation’s actions. They must ensure there are clear lines of accountability for the development and deployment of AI systems, and that ethical outcomes are a key performance indicator.

Responsible AI is a corporate governance issue. Any colleague on the board who tries to convince you that it starts and ends with the company’s tech team is kidding themselves. 

AI demands a lot of business calls and crucial decisions, and directors are the ones who make them. Therefore, they must ensure they know the essentials of responsible AI and how it feeds into the broader company strategy.