Getty Images

Gartner: Six key executive-CIO conversations to help AI succeed

The promise of artificial intelligence often fails to materialise in businesses, so how do you maximise your chance to succeed?

The promise of artificial intelligence (AI) often fails to materialise in our businesses – not because of technical limitations, but because of the executive leadership failing to achieve clarity on what AI is, and where and how it can add most value in their organisations.

CIOs have a powerful leadership opportunity to take their businesses through six executive conversations to address this gap.

1. What does AI mean in our business?

First, ensure everyone shares a common understanding and language about what AI is in their business context. Bring it down to simple use cases and examples. Talk about three AI categories: systems that behave like humans, such as chatbots; systems that automate humans out of the loop; and systems that generate next-level insight.

2. How does AI work with our workforce to achieve results?

Discuss how AI will work with your employees – does it replace them, help them work better, or work alongside them? All three can work well, but the shape of value and risk is different for each. Get executives used to categorising AI opportunities like this. 

3. How transparent is AI?

Consider how much you need to understand how AI is doing its job. Businesses must avoid unhelpful, unintended, maybe even dangerous or illegal bias in algorithms. Imagine an intelligent marketing algorithm that was shown to accidentally discriminate based on race or gender. The need for transparency may cause us to choose one AI technique over another, even if it performs less well. Executives must develop an acute awareness of this issue.

4. Which AI-powered business opportunities should we pursue?

Based on the answers to the above three questions, executives should decide where to focus AI activities. Decisions here can use the three-part typology of AI mentioned in question one, mapped against the different domains of the internal supply chain and ecosystem. This ensures we are not AI “fashion victims”, but instead consider the most valuable artificial intelligence opportunities across our whole business. 

5. How much are we prepared to rely on AI?

Combining the results of questions two and three above, executives can make high-level decisions and commission policies on how much AI-enabled automation is desirable and how transparent AI needs to be in different parts of the business. For example, a business might be comfortable with a completely automated “black box” that flags possible fraudulent transactions, but systems that make decisions about hiring might need to be much more “human in the loop” and transparent.

6. How will we manage and mitigate AI-related risk?

Despite making smart decisions about where to deploy which types of AI, there will be residual risk. The sixth executive conversation should cover types of risk, how to mitigate them and where accountability lies. Types of risk vary from injury or even loss of life, say from autonomous vehicles, through financial, brand and reputation risks. Developing a portfolio of techniques such as hedging with insurance and creating radical transparency with stakeholders is essential.

Gartner analysts will further explore key artificial intelligence considerations for executive leaders at Gartner IT Symposium/Xpo 2022, which will take place from 7-10 November 2022 in Barcelona, Spain.

Read more about AI in business

  • Bias cannot be stripped from artificial intelligence systems, but organisations can work to mitigate it.
  • Critics say you can’t trust AI yet with diversity and inclusion in hiring – or the elaborate audits meant to root out AI bias. Hiring software vendors beg to differ.

Read more on Artificial intelligence, automation and robotics

Data Center
Data Management