Contrary to popular belief, none of the major applicant tracking systems (ATS) currently on the market deploy AI in any meaningful way. That’s not to say the AI revolution isn’t coming, it just hasn’t landed yet. Given the lengthy procurement and implementation cycles inside major corporations, it will likely be a few years before we see AI deployed at scale, especially for recruiting. Most organizations will prioritize implementing AI in the parts of the organization that touch the customer or generate sales, in order to maximize revenue. HR is a cost center, and while the deployment of AI will help to reduce costs and maximize output, these functions are always lower priorities for investment dollars.
The pace of legislative change and the regulatory environment is a further hurdle to large scale deployment of AI: hiring practices are covered by federal and state laws, and there are significant burdens of proof required to demonstrate that any new technology (or algorithm) doesn’t infringe or contravene existing legislation in place to protect workers.
In this context I expect that the emergence of AI in recruiting will take place in waves, beginning at the operational and transactional level, before expanding into more complex functions in the hiring journey:
1. The Automation And Elevation Of Tactical And Transactional Tasks
It is helpful to think about AI (especially in the context of recruiting) in terms of a three-stage evolutionary process: even many of the legacy ATS systems currently in use have some minimal automation built into them. Some newer ATS systems deploy algorithms, but none yet have deployed Artificial Intelligence.
Perhaps ironically, it is this basic automation, and to some extent the matching algorithms, that are causing much of the discourse on LinkedIn around the need for job seekers to find a way to “beat the ATS.” As I have previously documented, all the major ATS systems currently on the market require so much recruiter intervention that it’s essentially an analogue process. There is nothing to “beat” because it is still a human making the decision.
Future-state, AI will be able to solve one of the thorniest issues facing recruiting teams today, and in doing so will actually make things better, not worse, for job seekers. On average, 75% of job applicants do not meet the minimum criteria for the role, and as such are considered “not qualified” and usually hear nothing back. This is of course a source of huge frustration to job seekers because they believe they are qualified, hence submitting their application in the first place.
Deploying AI will enable recruiting teams to solve this problem at scale: the key will be enabling real-time conversations with job seekers, helping them understand why they’re not qualified, in language they can understand and relate to. Furthermore I foresee AI being able to redirect job seekers to roles, or organizations, that are more commensurate with the experience and skills that they bring. AI will enable real-time feedback for a very large constituency of job seekers that are typically ignored in todays’ job market (recruiting teams are not funded, nor structured, to offer feedback to this audience).
Additionally AI will significantly expedite the many and varied operational elements that need to work together as candidates progress through the hiring journey (eg scheduling interviews, chasing feedback), meaning that hiring timelines will be significantly shortened. Removing this administrative overhead will also free up the human recruiters to spend more time with candidates and hiring managers, thereby enabling a better experience for everyone.
2. Unlocking Insights From Large Data Sets
The ugly truth about HR functions, even in large, well-funded, corporate environments, is that there is still an awful lot of guesswork. Deploying AI across talent organizations will finally enable true predictive analytics that will better inform hiring strategies. Traditionally hiring has placed an outsized emphasis on “pedigree”, because in the absence of actual data, this has been considered the next best thing. Recruiting teams often work with target lists of desired employers, which are small subsets of the broader industry or competitive set, because “folks from Starbucks often do well at Apple.”
In the future AI will be able to map out not just which previous employers better equip new hires to be successful in their new role, but which specific skills folks tend to have gained by working at company X or in industry Y. We will be able to move away from the rigid “four-year college degree and five years’ experience of X” to a world where AI will tell us that folks that have done “X” typically have the deepest set of “Y” skills, which is the best predictor of success at role “Z.”
Furthermore, just as the “single view of the customer” is the holy grail for marketing departments, AI will be able to unlock a single view of the candidate. This will allow recruiters to have much more personalized, and precise, interactions with talent, and will enable long-term relationship building. CRM for your career, basically.
3. Decision Augmentation
Perhaps the most controversial use-case for AI in hiring will be helping unlock better decision making, which is why it will take us longer to get there (not least because legislation will need to keep pace with the advances in technology). The best analogy I have here is the likely future deployment of AI in medicine: it is not inconceivable that AI will help radiologists arrive at quicker, better, more accurate diagnoses by acting as a co-pilot when they’re reading x-rays and scans. I foresee a similar application in hiring: AI will be able to observe interviews, will prompt interviewers to ask better questions, and will be better at interpreting answers than humans are. Furthermore AI will be able to spot potential instances of bias, and even escalate potential risks such as flagging micro-aggressions.
As I have previously written, a very large source of frustration for job-seekers is the lack of feedback from the interview process, largely driven by the risk-aversion of compliance and legal teams. Deploying AI into the interview process will not only allow the collation and curation of actionable, objective, feedback, but it will be able to run the feedback against all applicable legislation to eliminate any potential legal risk for the company.
Although we’re not there yet, I am already looking forward to a world where:
- Everyone that applies to a job has a meaningful and substantive interaction, and gets instant clarity on whether they’re a potential match, and what next steps and timelines will likely be. Everyone else will have a clear understanding of why they’re not a match, and some guidance as to where to head next.
- Administration and busywork is entirely removed from the value chain.
- Powerful predictive analytics significantly broaden potential talent pools, and unlock more opportunity for more people.
- Subjectivity and bias are entirely eliminated from the selection process.
- All interviewees get objective, actionable feedback.
- The right person gets hired into the right job 95%+ of the time.
When AI delivers these outcomes, everyone’s experience will improve.
