Mitigate AI Risks with Policies

Author: Victoria Kelleher

artificial-intelligence-digital-hand.jpg

Advancements in artificial intelligence are evolving rapidly, with applications that alter internal business protocols and customer experience alike. However, like any new technology, the vast potential of AI may outpace the ability of some employers to define policies that guide its use. In fact, even though AI has been a priority across various top businesses for some time now, a new survey from Brightmine found that only 34% of employers in the US have developed guidelines, principles or a policy to govern its use.

Regardless of whether there is a defined policy, most employers are at least experimenting with the use of AI in the HR department. Although only 20% have developed formal procedures for using AI in HR, an additional 60% of employers are informally experimenting with how it can be applied in HR. In addition, 78% of HR representatives who responded to the survey reported that that they have used AI in some way within their own workflows. 

It is no wonder that so many employees and departments alike are experimenting with how to incorporate AI into their work, considering its wide-reaching potential to enhance the efficiency of business processes and streamline the workflows of employees. HR professionals responding to the survey reported a variety of valuable applications for AI within the HR department, from streamlining administrative tasks to automating data and analytics (see Chart 1). 

Chart 1: Areas of HR that May Receive the Most Value From AI

The Risks of Unregulated AI

However, unregulated use of AI may pose risks if employees are not given clear guidelines to follow. Employers who have established formal AI policies have primarily done so either to prevent unethical use of AI among employees or to protect confidential company or client information.

Despite the optimism about how AI could potentially make things easier or more effective in the HR space, respondents also expressed concerns about the risks of improper use of AI. They were especially concerned about the risk of AI causing biased practices in HR areas and about the risk it poses for legal compliance or data security (see Chart 2). 

Chart 2: HR Perspectives on the Risks Around AI Adoption

Defining an AI Policy

AI may raise some risks when it is used across an organization without precautions, but banning its use altogether has its own risks. Employers who do not adapt to advancements in AI will not be able to reap the benefits of this technology. They will miss out on the efficiencies it can offer, and their performance may fall behind that of competitors who have embraced AI's potential. 

To take advantage of AI while mitigating risks, organizations should develop a formal AI policy to regulate the use of AI among employees. Employers should make sure to address several points within this policy, including:

  • How AI intersects with other subjects like cybersecurity and data privacy;
  • What type of AI can be used and for which purposes, and;
  • The risk areas around AI ethics and employee conduct.

The process of creating an AI policy should incorporate multiple stakeholders, and there should be an emphasis on understanding the functions of AI, the risks associated with its use, and relevant considerations for legal compliance. Employers may benefit from creating formal training resources for the working group and including a legal advisor in the process. Taking these precautions will ensure that the resulting AI policy is comprehensive and covers all the bases necessary to clearly define safe and ethical use of AI across the business.

Takeaway

AI undoubtedly has tremendous potential to boost worker productivity and efficiency within HR and beyond. It is safe to assume that employees will experiment with how AI can enhance their workflows whether an employer addresses its use in company policies or not. To mitigate the risks associated with unregulated use of AI, it is imperative that employers set clear guidelines for permitted use cases. This way, employees are assured in their ability to use this technology for all its benefits without opening the organization up to avoidable risks.