Information contained in this publication is intended for informational purposes only and does not constitute legal advice or opinion, nor is it a substitute for the professional judgment of an attorney.
How should a company intelligently adopt employment-focused artificial intelligence, or AI tools?
Employment-focused, artificial intelligence-driven tools promise – and deliver – significant operational efficiencies, cost savings, and enhanced workforce and workplace outcomes. These include tools related to recruiting, talent assessment, communication, disability accommodations, and safety. However, the inner workings of these emerging tools and technologies cannot be simply explained. Further, there is no clarity on how the use of AI will be regulated by Congress and the courts, particularly in light of heightened public concern that AI algorithms are capable of discriminatory bias. Forward-thinking corporations can take a pragmatic approach in adopting AI processes.
- Consider whether you need AI to begin with. Even if a tool promises to address your entire business, start with a specific use case – one key business activity you want to enhance – and make sure that the anticipated improvements to that process will be critical to your business. Then make sure that the AI tool under consideration will achieve that goal better than any other.
- Investigate multiple vendors and tools, and ask detailed questions on four issues:
- How the product was tested to ensure the absence of legally significant bias
- What warranties, indemnity or other support the vendor will provide if litigation results
- Whether the vendor has qualified testifying experts on hand to explain and defend their product
- Whether the vendor can provide confidential references of other businesses that have used the tool for that same purpose
And once you’ve adopted the tool … conduct or obtain initial testing to ensure that legal standards are being met, work with the vendor on enhancements, document all your efforts to be a good corporate citizen, and review your results to make sure the test is leading to the desired business outcomes without creating a disparate impact.