Although Artificial Intelligence (AI) has been around in one form or another for the past 30 years, with the advent of the web and ability to electronically link large datasets, AI is coming on its own. The AI transformation from work to daily life is coming on strong and fast, but not fast enough to immediately eliminate a number of jobs.
For the most part, large organizations with a healthy IT budget are implementing aspects of AI, primarily in the recruitment sphere. Employers have to ask when using or implementing new applicant tracking systems (ATS) whether some sort of AI tool is being used and what data-set it is deriving its decision-making process on. Two cases that employers need to be aware of that concern AI in the recruitment sphere include the Workday lawsuit in the federal 9th Circuit in which an applicant alleges that the Workday ATS used an AI tool that screened out the candidate illegally, and the EEOC settlement of an age discrimination case in which the company iTutor, a Chinese company recruiting English language instructors, had its AI tool screening out anyone essentially over 55 years of age.
Another company is using AI to summarize recorded conversations between doctors and patients. This is the tricky part. Employees need to double check the AI, because it does make mistakes. In this case, doctors need to diligently double check since their careers and reputation are on the line. In other areas of business, such as operations or call centers, where AI could be used to summarize customer-service calls, there is more of a risk that it won’t be diligently double checked.
Yet AI training is not a priority of employers yet. A recent survey by Randstad shows that only 13% of employees have been offered any AI training in the past year. Without training, employees may use these tools inappropriately, including putting confidential information into the tool, which cannot be deleted, or trust the tool not to make mistakes. That alone is a mistake. One New York Times reporter had a “chat” with ChatGPT and after two hours the tool threatened the reporter’s life. So, AI isn’t totally there yet.
HR needs to be on the forefront of this transformative technology. It’s almost like being at the forefront of the computer evolution back in the 1970s and 1980s and rise of the Internet in the 1990s. The impact will be great on employers and employees.
First, employers need to determine how and when it will be used, from coding to customer service to the manufacturing floor to health care to HR. With a labor shortage expected to continue for a long time, AI could bridge that gap, like other technologies in the past. Once leadership determines how AI will initially impact the organization, identify the persons in the organization that need to be initially trained on it.
Next, HR with legal counsel, must craft policies and practices that will lead to compliance and protection of confidentiality and trade secrets. For example, any policy could ban ChatGPT because there is no control over what is put into the dataset. In addition, legal counsel should inventory the laws that impact AI usage from federal to state to local. Currently, NYC is on the forefront, but other jurisdictions will follow, and similar to paid leave and pay transparency, the laws will be all over the place from high regulatory environment to none at all. At that point, HR has to recommend what standard they will use to monitor AI usage in the organization.
Last, employers need to identify experts (data scientists) who can assist with the validation and review of the datasets being used or proposed to being used by the AI tools supplier. This aspect is extremely important and most of HR will likely be ignorant of the technical aspect of AI. It shouldn’t be implemented in any form until the dataset is cleared for use. For example, Amazon used an internal dataset of IT resumes to test a new recruiting AI tool. Turned out that mostly white males were selected – makes sense when reviewing the composition of the IT workforce, which is mostly white males.
AI is here and not going away. Therefore, HR must be prepared and ready to implement appropriately. Although your organization may not be using it, AI will permeate the tools HR uses. Be prepared for the future.
If you need assistance with evaluation experts, ASE may be able to assist. Contact Anthony Kaylin at [email protected] for more information.
Source: Wall Street Journal 9/8/23