Practical implementation and governance: building an AI-ready business

If you haven’t read our AI-related blogs before – what you’ll find here are ‘the reality of change’ and ‘shifting mindsets away from fear of AI’ – one thing is clear: AI is becoming an increasingly important tool.

With this in mind, business leaders must consider how best to prepare their organizations and people for these changes, so that these tools can be used successfully.

To better understand ethical AI leadership and technology change management, we hosting webinars with HR Director Lizzy Barry, HR Consulting Director Dan Grace and Senior L&D Advisor Vanessa Aradia.

In this blog, I’ve summarized some of their conversations about how you can prepare your business for AI.

Create AI policies and guidelines

During the webinar, Lizzy discussed the importance of having strong AI governance, saying: “We previously covered lawsuits that arise when applicants are unsuccessful due to AI bias.

“This is a clear example of some of the risks we face in using AI without implementing checks and balances – this is no small matter or small legal issue.

“Ethically it is also not right, because what can be changed is people’s lives with this technology.

“We all have a moral and legal obligation to ensure we use these tools effectively.

“Some of the things we’ve done at IRIS, guided by best practices and upcoming legislation in different locations, is ensure we implement important data privacy, data security and AI use policies across the organization.

“Is it clear to everyone in the business what AI tools staff can use in the workplace, and what type of information they can input into those tools?”

“Are we teaching our employees how to use AI so they understand that not everything it produces is correct?

“I’m sure many of you have heard of it a situation where a law firm decides the AI ​​can write multiple summaries and do background research on some of the cases they are preparing, and the AI ​​generates all the citations and statutes it references.

“No one checked the information before it was published, and it was a huge shame that it made international news, because it highlighted not only the failure of AI to do anything right, but ultimately also the failure of the people who use it.”

AI literacy

Lizzy also shared some insights into IRIS’s AI literacy program, detailing the areas we cover, such as:

• How to rate AI-generated content
• How to fact check the information
• Review images and videos that AI can generate
• What information is safe to share with AI

He added: “We must ensure that we manage all these risks in the same way as we manage other areas such as finance and data protection.

“We need tight security, and to ensure that people understand how to keep businesses safe.

“Employees also need to know where to go if they have questions or concerns.”

Practical steps to prepare your business for AI

When preparing your business for AI, Vanessa shares some practical advice:

• Start small: identify high-volume, low-value processes that are ideal for automation
• Implement the SAFER framework: sustainability, accountability, fairness, explanation and resilience
• Cross-functional governance is key: An AI strategy must involve HR, IT, compliance, and operations
• Build explainability into every AI application: users must understand what the system is doing and why.

Vanessa explained further, saying: “Government is the guardrail for trust.

“A transparent audit trail and clear ownership model protects the business and people behind the data.”

“If your AI can’t explain itself, it won’t make decisions.

“At IRIS, we have created guidance for teams on what data is safe to share, when to involve human oversight, and how to validate output.

“Such clarity protects trust because people know where they stand, and trust is the foundation of any successful technology launch.”

Avoid AI bias

AI bias is becoming an increasingly common problem, which Dan touched on in the webinar, saying: “The uncomfortable truth about AI is that because systems are trained on historical data, they will have historical biases that are present in everyday media, especially when available on the internet.

“The Internet is a repository of unprotected information.

“There are many unbiased opinions – in fact, they are usually very biased.

“In terms of practical mitigation strategies for using AI, first of all, I recommend that if you want to use an AI tool, always pay attention to pre-implementation due diligence, which can include requesting a bias audit from the vendor.

“A bias audit is different from their marketing materials.

“This should be real audit results, covering the source of the training data, how they verified it, and the demographics of the data – this will help you establish bias in the tool.

“Second, you must use diverse representative training data, applying fairness and awareness algorithms.

“If you don’t have a technical mind, the vendor should be able to help with that, or if you have an in-house software engineering team, they can provide support as well.

“With AI, the non-negotiable is that you have to be constantly monitoring.”

Webinar: Beyond the hype – ethical AI leadership and change management for HR and payroll

If you want more guidance, good news! Our no-nonsense practical sessions are now available to watch on demand.

In this webinar, you will learn how to:

• Turn employee fears into technology adoption with ethical leadership
• Build AI-enabled organizations to amplify human potential
• Create a practical governance framework for data privacy, security, and corporate policies
• Develop employee training and AI literacy programs
• Change your HR function from reactive to strategic

This session will give you the framework and tools to confidently lead through AI transformation, without losing the human touch that makes your organization special.