Anyone watching the tech sector knows that Artificial Intelligence (AI) is scheduled to explode over the next several years.

The AI sphere will go from its 2018 market value, $21.46 billion, to a value of $190 billion by 2025 – it’s scheduled to increase nine times in just seven years. One AI capability in particular that’s changing possibilities in many different industries is facial recognition technology – that market will grow to $9.58 billion by 2022 (growing at a CAGR of 21.3 percent since 2016).

The ways we relate to businesses, government and even each other will likely be reshaped. AI’s advent is often referred to as a “revolution” for good reason.

For example, facial recognition technology is swiftly finding applications in device security, public safety, finance and retail. Devices such as phones and tablets are being developed to unlock by scanning a person’s facial features and certain stores may soon allow customers to “pay” with a face-scan transaction.

AI will also, within the next one to two years, be able to perform conversation-based tasks. Companies are already beginning to use AI to automatically sort through hundreds (sometimes thousands) of applications, narrowing the pool without the need for hours of tedious human labor.

Clearly, AI’s potential to enhance our lives is nearly limitless. But we have a responsibility to make sure it functions ethically and without some of the flaws that plague humans. The problem is that a tech workforce in a state of diversity crisis is programming some of those same biases into the AI systems it generates. These systems then become forces that perpetuate rather than alleviate many of the difficulties women and minorities already face.

It’s notable that Amazon scrapped an AI hiring software project, which ran from 2015 to 2017, when developers realized the machine learning algorithm was discriminating against female candidates. This occurred because of the preponderance of male applicants versus females.

For an industry that produces products which affect so many aspects of our lives, the lack of diversity in the tech sphere (and its AI subcategory) is truly worrisome. Women make up 15 percent of AI research staff at Facebook and only 10 percent at Google. Only 18 percent of speakers at leading AI conferences are female, and 80 percent of AI professors are male. Just 2.5 percent of Google’s workforce is black as well. And despite a growing awareness of this issue within the industry, it doesn’t seem to be changing fast enough. Currently, 52 percent of women consider leaving their jobs within the tech industry.

Any tech company serious about ensuring its success and relevance for the future – and any AI developer serious about producing software that serves everyone equally – needs to consider workforce diversity. And it’s not just a “nice to have” thing or a chore to undertake – diversity entails serious benefits for companies that commit to it. According to a study by the Harvard Business Review, companies that are more diverse are 45 percent more likely to report that their firm’s market share grew over the previous year and 70 percent more likely to report that it captured a new market.

The reason? A more diverse company is much more capable of innovation, attracts better talent and better fosters an environment where everyone feels their contributions are important.

The Harvard Business Review additionally found that when at least one member of a business’s team shared a client’s ethnicity, the team as a whole was 152 percent more likely to understand that client. Workforce diversity is the only way to truly stay relevant in a rapidly diversifying U.S. and world.

So how can HR in tech companies – and especially companies that use and develop AI – get serious about incorporating and fostering greater diversity? Consider the following suggestions:

• Publish employee compensation levels, including bonuses and equity, across all roles and categories, including executives. Show the breakdown across race and gender.

• Maximize diversity in hiring practices: look beyond elite universities for candidates, connect with community and nonprofit organizations designed to promote diverse hires and develop a more equitable focus on under-represented groups.

• Tie executive incentive structures to increases in hiring and retention of under-represented groups.

• Track and publicize, at all times, where and for what purposes AI is used at your company.

• Test any and all AI software for performance around sensitive issues relating to bias and discrimination. Take these measurements regularly.

• Conduct a thorough risk assessment, given issues of bias and fairness, as to whether certain applications should be designed at all (just because you can, doesn’t mean you should).

The lack of diversity in the tech sphere, especially where rapidly expanding AI is concerned, has the potential to carry forward inequity and injustice that we know we need to leave behind. Become a leader in your field by choosing to be a part of this solution and allowing technology to truly contribute to a better world for us all.

Vladimir Sidorenko is Founder and CEO of Performia CIS, an international personnel management consulting company. Created in 2001, they specialize in effective solutions to personnel problems and technology for hiring productive employees and contributing to higher profits for companies. Performia International is headquartered in Stockholm, Sweden and Performia CIS (Commonwealth of Independent States) is located is Moscow, Russia.