UK 
United States
AI (or artificial intelligence) in recruitment is a current hot topic. Go to any recruitment or HR conference and the usual talking heads will be proselytizing AI’s revolutionary impact for our industry.
I think you'll agree- this is a fairly broad prediction. The truth is- it is really hard to find out exactly how these experts think AI will actually improve recruitment. Try to find some specific, tangible examples of really revolutionary benefits achieved... it's actually quite difficult.
And that's because there's a secret that the promoters of AI are not telling you. Recruitment AI either:
a) already exists in the form of automation.
OR
b) has a narrow scope for application in recruitment.
Let me elaborate…
The difference between automation and AI
Automation exists in recruitment software right now. It is all of the actions that your recruitment software takes without you having to tell it specifically what to do. It does these actions because it was programmed and designed to. Some examples include:
Mail Merging (Email / SMS / Documents)
Merge Fields pull specified fields out of your system and insert them automatically into email, SMS or document templates. So, when you email 200 candidates simultaneously, you don't need to enter 200 first names. It's clever- it's useful- it's not AI. It's automation.
Automated Job Alerts
When candidates sign up for your job alerts, you don't need to email each and every candidate with the details of every new job. They receive updates automatically. Really useful feature- 100% not AI. It's automation.
CV Parsing
When you upload a CV to your recruitment software, CV parsing technology automatically takes information out of that CV and populates contact information and skills into the relevant fields in your recruitment software. A huge time saver- really useful and 100% automation. Not A.I.
Chatbots
Not as widely used as the previous 3 features- Chatbots mimic human conversations on either your careers site or social media. They are programmed with a limited number of responses to a limited number of scenarios. Definitely automation.
SO…
Most of what people refer to as AI in recruitment is actually automation. Automation is great! Really- it saves recruiters all over the world hundreds of thousands of hours. It's also mind-bogglingly clever. But it's not A.I. Here's why:
Artificial Intelligence is defined as:
Any device that perceives its environment and takes actions that maximize its chance of success at some goal. Colloquially, the term "artificial intelligence" is applied when a machine mimics "cognitive" functions that humans associate with other human minds, such as "learning" and "problem solving".
Automation can't solve problems or learn. So that's the difference between A.I. and automation.
Some Practical Applications of AI in Recruitment
There are some ways recruiters can use AI in order to improve candidate experience and reduce cost of hire, but these are limited to specific areas of recruitment. Here are some practical examples of how you could incorporate AI into your recruitment processes.
Chatbots *Not a typo*
Yeah I know I said chatbots are being used as part of automated recruitment features right now. However, Chatbots have to potential to perform cognitive functions- which would make them AI. Some advanced chatbots are already learning how to increase their range of responses by analysing conversational trends. These chatbots could replace human agents on the end of live webchats.
Telephone interviewing candidates
Voice recognition technology is getting really advanced. Some of the most advanced systems are already difficult to tell apart from human counterparts. AI screening could conceivably learn what the best answers are to initial screening questions and rank candidates accordingly.
Translation
You need to hire a French speaking social media expert. But you don't speak French. In the past this could have been an issue, but real time translation is getting really very good. You could potentially interview a candidate- speaking English- they could reply in French and you could both understand each other perfectly. Technically translation is a cognitive function- so definitely counts as AI.
Why Your Job Is Safe...
In summary, AI could definitely improve on current recruitment processes. There are some useful features that can support and improve the way we hire- and more possibilities than mentioned here.
But we believe it still has limited applications- and website willrobotstakemyjob.com agrees. The site estimates there is a 0.55% risk of automation for Human Resource Managers. And while there's no corresponding data for an In House Recruiter, we could expect a similar result due to the similarities of the positions.
HR and by extension recruitment are "people" jobs. They are about building relationships- being sensitive, taking into account personal situations. It requires an extremely specialised set of skills that most humans struggle with- so how will artificial intelligence fair?
AI is unlikely to 'give a candidate a chance'. It will never hire the unexpected candidate that turns out to be a star performer. Will machines ever understand the importance of cultural fit? And, if so, will they be able to make a judgement on how well the candidate will fit into your company culture? These are intricate and fundamentally human decisions. AI has some way to go before it can understand them.
We can't see AI taking over any time soon.
Clodagh O'Reilly @cl
Cognitive recruitment tools now can take volumes of data no individual or collection of recruiters could absorb and retain, then rigorously make sense of it and subsequently offer meaningful advice on when and how to apply it.
Machines can learn to intelligently review resume submissions to identify the extent to which an individual has the skills required for a role. Which means, when you open a requisition and get a dozen applications, or a thousand, it can efficiently rank them all on the basis of skill fit. But that is just the beginning.
Consider what it will mean to have a machine learn the contents of your applicant tracking system (ATS). First, by ingesting information on all the applicants who have ever captured or parsed a resume/CV in your ATS. Then incorporating information on how they did in the process, associated assessments, ratings and hiring decisions. Based on its training, it can then draw conclusions on why and how individuals succeed or fail in your organisation’s recruitment process. Then it can analyse any new application you ever receive against those criteria for success and give you a score of the likelihood the individual will succeed in your hiring process and even organisation (depending on whether you have also supplied post-hire performance information for ingestion), instantaneously.
You might be looking at a computer screen that appears almost identical to your ATS screen, but the work going on behind it is far superior.
So you can open a req and have 100 applications in overnight, arrive in the morning to a fully ranked candidate pool accounting for each individuals (1) skill fit to the requirements of the role and and (2) likelihood of succeeding in your organisation.
And, if you feel that is too ‘scary,’ you can get a full explanation of where the data was derived from, the analysis that was undertaken to deem it relevant or otherwise, and so be able to understand any particular recommendation... and that's just the start.
Machines can also learn from your records and data available on job boards, if your role is likely to be easy or hard to fill, where you might find the best candidates and how long in advance hiring activity should commence. Also, effortlessly, a truly cognitive recruitment solution can help you understand how to attract the most desirable talent. Using social listening and analytics, it can discern your current employee opinion/brand reputation and compare that with competitors' brands and employee opinion. So you, the recruiter, have ready access to top tips on how to ‘sell’ your organisation to your top candidate, focusing on your organisation’s strengths and competitors weaknesses.
I am confident this sort of information will significantly accelerate recruiter decision-making and candidate management, resulting in a more efficient and positive recruitment experience for all. I believe these potential gains should be taken seriously for organisations that are competing for talent (aka most organisations).