As the debate around ethical artificial intelligence (AI) continues, a recent McKinsey & Company article warns business leaders that relying on company values only as your “compass” is not enough to navigate this new territory. In fact, it’s CEOs themselves who need to provide guidance, leading employees into the promised land of artificial intelligence and machine learning.

Beyond the numbers—like revenue and profits—McKinsey makes the case that other measures of success come into play in the age of digital transformation, including “making sound decisions that not only lead to the creation of value but also do no harm.” This latter set of metrics is where AI especially comes into the picture. 

AI is already delivering big for companies across all industries. In fact, according to McKinsey’s research, the number of companies implementing AI into their business doubled from 2017 to 2018 with “nearly all” companies achieving business value as a result. But the firm points out the need for “careful management to prevent unintentional but significant damage.” This includes of course any negative impact to your company’s brand reputation, but also extends to your employees, citizens, and society overall.

In this environment, it should come as no surprise that CEOs in particular are “under the spotlight” to strike the right balance between business results and corporate social responsibility. This extends beyond compliance and helping employees update skills. In fact, according to the McKinsey article, CEOs need to “dig deeper” into the impact of AI by challenging their data and analytics teams to constantly evaluate their actions. For the CEO’s part, asking the right questions at the right time is a key part of the process. 

To that end, the McKinsey article offers five key focus areas to help guide CEOs. Read more below on the potential impact they can have on your business:

  • Appropriate Data Acquisition: The pressure to perform better often requires new data sources and more of them. In the endless quest for more data—including from third-party sources—CEOs are advised to be “vigilant” in asking where data is being sourced and in understanding all of its implications, especially when customer data is at play.
  • Data Set Suitability: More accurate data sets lead to more accurate results. This is critical when it comes to leveraging AI to its full (and accurate) potential. But this requires CEOs to ask very specific and “granular” questions about the data sets being used, while keeping a broad perspective on all the potential populations a given data set should include.
  • Fairness of AI Outputs: Even with a sound ethical framework in place, AI bias can still occur, especially when historical data is being used. To ensure fairness throughout the process, CEOs are encouraged to design “fairness” into the process by properly defining it up front and applying metrics to measure it throughout the process.   
  • Regulatory Compliance and Engagement: The McKinsey article urges CEOs to not only ask questions about regulatory requirements and compliance but also promote collaboration between data science, regulatory, and legal teams with shared metrics. 
  • Explainability: Being able to clearly explain how AI is making predictions that help your business—while remaining ethical—is another area where CEOs are being called to task. According to the McKinsey article, business leaders should know their data analytics teams and initiatives well and strive to truly understand the data models they use when implementing AI.

To learn more, including how company values are part of the mix, read the complete McKinsey & Company article.