Business Today

AI can eliminate biases in hiring, but has limitations

Since AI gives us the ability to sort through massive amounts of data, it can help identify trends

twitter-logoGoutam Das | August 2, 2019 | Updated 22:18 IST
AI can eliminate biases in hiring, but has limitations

Artificial Intelligence (AI) is increasingly playing a crucial role in human resources - both in hiring and talent management. Nevertheless, people have started realising that exponential technologies have limitations but could work better with the right data models and nurturing.  

Ryan Ross, Managing Partner of Hogan Assessment Systems, a company with expertise in personality assessments, made an interesting presentation  on AI's role in talent acquisition at the People Matters TechHR 2019 conference in Gurgaon, on Friday. He later spoke to Business Today to highlight what AI can, and can't do. Here are some of the highlights:  

  1. AI affords us the ability to get massive amounts of historical data and find nuggets of truth and other key points that otherwise would take us, humans, years, he said. Since AI gives us the ability to sort through massive amounts of data, it can help identify trends. "The challenge is, do those trends make operational sense? If you evaluate all your leaders and they happen to have size 9 shoe, does that mean you only recruit people that have a size 9 foot? AI and machine learning will tell that is all you should do. So you have to add a contextual element to the data analytics," he added.  
  2. AI can help eliminate some of the biases and discrimination that show up in hiring and recruitment processes. "In some parts of the world, different races are excluded from the hiring process simply because the hiring manager doesn't like them," Ross said. AI may do better but the challenge is that modelling of historic practices could reinforce that discrimination. "You have to start with a blank slate so that the machine learning algorithms don't pick up historical biases," he said. Machines can be safer if algorithms are fed with proper criteria such as personality data. "Personality data is non-discriminatory. It is gender or age blind. People of different gender, age and colour score the same on personality-based assessments. If you are feeding valid predictors of performance into the algorithm, you are bypassing the bias a hiring manager may have," Ross explained.  
  3. AI also has a role in learning and development. The younger workforce wants targeted, on-demand learning instead of broad programmes. They prefer development modules around their strengths and challenges. AI can identify micro-learning for individuals quickly.
  4. What is it that AI can't do when it comes to human resources?  "It is not HR tech," said Ross. "It should be human HR. Put the human back into HR". His reasoning: employees want a real person answering questions rather than a chat bot. "We have assumed that because people use AI in their daily lives, that is what they want when it comes to very personal things such as employment, wages, performance discussions. We are overestimating AI's ability to replace a human being."

Also Read: India's top court asks central bank to confirm WhatsApp's compliance with local data rules

Also Read: IBM fired 1 lakh older employees to look 'cool and trendy', claims lawsuit

Also Read: India behind UK, France as 7th largest economy in the world in 2018: World Bank

  • Print

A    A   A