Avoiding tech pitfalls in human resources and hiring.
Throughout the Human Resources(HR) industry the implementation of AI tools has been leveraged with success to moderate the administrative load. The immediate win for AI enabled teams is freeing up more hours to engage deeper with people and develop stronger organizations.
While it can be tempting to rapidly expand upon these initial successes, many have discovered that these tools, when left unmanaged and unsupported, have the propensity to affirm human biases. Try as we might to overcome biases, they are an unfortunate affliction of the human condition. Every day our brain attempts to keep us safe and informed by finding patterns in the human behavior around us. These patterns and predictions are building the blocks of subconscious bias. When not addressed directly these biases can influence every aspect of our work, to include developing algorithms for AI. As a result we’ve observed these potential tripping points in AI application in HR:
- AI resume reviews: When criteria is set for resume screening tools there is of course a long list of role specific terms chosen, but certain choices of common terms can also act as a vehicle for bias. This was experienced by Amazon first-hand in the early days of AI application. Over a number of years they realized that they were seeing a higher volume of male engineering applicants rather than female due to the male tendency to use the verbs “executed” and “captured” more frequently on resumes.
- AI Interview transcription: The video interviewing process has been a boon to busy professionals on both ends of the hiring process. It works by allowing candidates and hiring managers to engage the process within their individual time constraints. AI led transcription and scoring tools is where trouble starts. Body language and tone already exists in the realm of nuanced communication, with lots of cultural and individual variations. AI trying to apply a broad formula of behavioral expectations to candidates, and accurately score interactions accordingly, has resulted in challenges. This particular scenario was made real in 2020 when a MAC Cosmetics employee sued when she wasn’t rehired after a company furlough as she failed to pass the AI body language screener.
- AI performance data: Some teams are utilizing AI to parse through immense quantities of data to discover trends in employee productivity and wellness. While this data can be helpful in shaping success internally, applying it as an archetype for hiring can result in homogeneous teams.
- AI aversions: Some of the early adoptions of untested tools have actually resulted in negative associations. A survey conducted by Harvard Business Review concluded that only 31% of respondents were comfortable with AI’s influence in the workforce, with the rest being neutral or alarmed by its use. This in and of itself is an obstacle to overcome as AI continues to improve in quality, and drive efficiency.
As the technological landscape continues to evolve, HR teams are positioned at the heart of these changes, driving innovation, employee satisfaction, and serving as a guardrail against technological shortcomings. For HR professionals, this presents exciting new opportunities, as well as challenges to overcome. We at the Frank Group are prepared to come alongside organizations experiencing growing pains, to support and build teams that will allow them to thrive long term.
