If you are hiring in China these days, it’s hard to not stumble over some sort of AI service offering. We’ve tracked it since the early chatbots and Natural Language Parsing (NLP) apps. Some recent articles point to the evolving and important trend of AI in recruitment.
- The legal case that forced Hirevue to drop its emotion-analytic video interview tech
- Rocket-Hire’s research suggests that “81% of the investment in predictive hiring tools is going to AI solutions”.
Coupled with the recent announcement of Tsinghua University’s ‘trillion parameter’ AI-powered ‘student’, we took a fresh dive into the state of AI in recruitment in China.
Here are 5 takeaways to help keep you up-to-date:
1. Booming Investment
Rivers of it. Pouring in since 2018, the taps still look fully ‘on’. Big jobs boards have already built or bought AI tech. Mega companies like Ping An have jumped in too.
2. Hiring platforms + Niche players
51job, Zhilian Zhaopin, etc all have AI-powered recruitment offerings and ambitions to cover all parts of the hiring process. There are plenty of niche AI talent assessment companies too – like HRTPS (linked to 51job), å£¹é¢, å¾—è´¤, å¤šé¢, and Hong Kong’s -Neufast - and business appears to be expanding rapidly.
3. The AI interview is no. 1
You can find “AI” in a broader range of offerings now: resume screening, chatbots for ‘engagement’ and job discovery, minimum qualification screening interviews (mostly text-based, but some with avatar interviewers), and interview scheduling. Many also interface with WeChat mini-apps.
The biggest target for AI solutions in recruitment is definitely the interview. Related methods and assessment outputs are evolving. From 5 interview questions, å£¹é¢ report a ‘professional ability score’, and ‘career risk’; similarly, å¤šé¢s interview sets out to ‘automatically analyse’ cognitive level, personality, and motivation. å¾—è´¤, who use NLP and facial analysis, adopt ‘structured interview techniques’ and aim to fully simulate human interviewer behaviour. HRTPS and Neufast produce competency reports from AI video interviews.
4. Theres plenty to be positive about
- Efficiency bonus – AI offers to automate more and more of the hiring process. The Ping An case includes some compelling efficiency improvement numbers.
- Matching human interviewer capabilities – Given the known bias and relatively low validity of many human interviews, replicating current validity with AI would be a more than acceptable starting point. It might even encourage humans to raise their game.
- Technical proficiency – There are now some very skilled AI technicians behind the current offerings. For the first time, we are seeing ISO standards being added to marketing blurb. We also know that business psychologists are becoming more involved in design and delivery.
- Hot competition – The AI game is still very much alive. That’s good news for customers. Large companies appear to be exploring options and looking for “all-in-one” AI hiring solutions.
- Moving into the organisation – A few offerings are looking beyond recruiting to include training and performance, linking the whole talent cycle and potentially releasing a predictive validity dividend.
5. There are obvious concerns for talent assessment
- Buyer beware – The main AI hooks are efficient and have a modern brand feel. Nothing wrong with either. Effectiveness (actually producing better hiring decisions) seems to be overlooked in seller propositions and probably from purchasing decisions.
- Validity? – Even if HR was looking for more effective assessments, beyond vague marketing blurb, there is still very little evidence of improved validity of the measurement, compared to existing approaches. Chatting to our friends at SHL China, this is their primary concern about current AI applications. Equally, given the current legal context, there is no likelihood of a Hirevue-like takedown in China, yet.
- Transparency – Without the solid professional and technical knowledge, access to suitable data, and conducting ongoing research, bias could be rampant and worse than ever.
- Privacy – AI tools trawl personal data… a reasonable concern for most candidates in China as elsewhere.
- Candidates learn fast– candidates in hyper-competitive economies like China are already learning to ride the new AI-interview pony. Coaching and cheating probably abound in equal measure. (An example from South Korea, in 2020.)
Overall, the progress made in just a few years is impressive. Continuing investment, robust competition, stronger IT capacity and technical standards, and more involvement by business psychologists all point to a promising future.
Will HR adapt so it can lead and manage these new initiatives? Ask me another one.
The post \"Chinas AI recruitment lovefest\" was first published by James Morley-Kirk here https://www.linkedin.com/pulse/chinas-ai-recruitment-lovefest-james-morley-kirk/
About James Morley-Kirk
Business Psychologist specialized in talent assessment solutions for individuals, teams, and companies in China, Malaysia, and Asia. Director at China Select è¡¡æ°. Executive Coach with eCoachPro.