AI Disruption of human capital management

Artificial Intelligence (AI) is a hot topic that has taken over conversations globally and is being widely debated – helpful or harmful? John Burke, Head of  Business & Professional Services in Citi Commercial Bank, has laid out a few important questions about how AI is going to affect the Human Capital Management (HCM) subsector.  Rob Garlick, Head of Innovation, Technology and the Future of Work at Citi Global Insights, provides answers on how AI is already impacting HCM and what the future holds.

1. What components of HCM are most likely to be disrupted by AI in the next 5 years? 

HCM is one of the largest technology sectors in the world and AI can bring substantial changes across the Learning & Development (L&D), Talent Acquisition and Talent Management subsegments. Many of these changes are already underway and are growing quickly, so while AI only made-up 15% of the HR tech market in 2023 it is forecasted to become 25% of the market by 2027.^[1] One factor contributing to this is that the AI toolkit just received a significant upgrade with transformer AI technologies - improving language comprehension and better integrating multi-modal features such as computer vision and sound used in video interviews.

L&D is a large area where outcomes are enhanced by more personalized AI Adaptive Learning Systems, with gamification also increasing engagement levels. A new HCM tool we are excited about is Talent Intelligence Platforms (TIP) that will help companies upskill their employees. Everyone agrees that there is a big need to do so. For example, the WEF estimates 44% of worker skills will be disrupted in the next 5 years, partly due to AI and automation^[2] Employees care greatly about upskilling, but all agreeing is not enough. Corporate spending on workforce education has been in decline in real terms over the last two decades. Fortunately, there is an AI solution to this AI problem, in the form of TIPs which use AI models to help infer, inventory, and suggest upskilling opportunities to employers and employees.

15% of HR budgets are spent on recruitment.^[3] According to JobScan, almost all (97.4%)^[4] of the Fortune 500 companies use some form of AI in their hiring process, but we expect usage will expand within these firms and to smaller firms, partly due to new AI transformer technologies. 

2. Which AI tools can be most valuable for HCM firms? 

Good managers have always helped upskill employees, but often there are business constraints in addition to budget limitations. TIPs can help. Companies are the ones investing in new growth areas and have the best foresight on what they need from future talent. TIPs can help audit the skills of the current workforce and suggest who would benefit from upskilling opportunities rather than firing some and hiring others. It is often more cost effective and efficient to migrate internal talent.

TIPs often use external compensation and skill trend data to help inform managers, but in addition can create internal talent marketplaces where workers can see new opportunities – half of workers say they don’t see internal opportunities and only 17% of employees feel their manager helps them apply for new jobs.^[5] This talent hording issue is not new, but it does create attrition. And while TIPs are only a tool, they can help all parties try and create a culture of continual upskilling to keep up with our ever-changing world.

3. What HCM functions is AI not ready for? 

There are many sensitive and personal HR areas where AI is inappropriate and there are many other areas when it is imperative that humans make the final decision such as hiring or firing, but AI can be a support tool. Take hiring for instance. In our latest report we highlight that AI can help in over 90% of a typical hiring process. ^[6] Research suggests AI can already better grade talent in the initial application sifting process if the volumes are high.^[7] Similarly, AI can judge personality traits more accurately than humans via psychometric testing. And while one-way video interviews are new they can assess soft skills – HireVue for example has hosted over 70m interviews and can assess 20 different competencies.^[8] When, as we discovered, there can be 4,000 applications for one role, AI can really help HR narrow the field to the ‘best of the best’ to take to human interview and selection.

4. What do firms need to do in order to prepare for and take advantage of AI? 

Many firms already are taking advantage of AI in HCM, so it depends where firms are in their journey. But given AI is moving at pace, a good place to start to understand where leading firms are and the innovation taking place is to read our two recent reports on the subject, the first on TIPs and the second on hiring. There are suppliers included in this, some of whom provide a full suite, from hire to retire, and will integrate with firms existing data and IT infrastructure. While data can be key to extract useful insights, outcomes and a good ROI, there are lots of developments on the data side and AI transformer models are game changing.

5. How will AI benefit HCM efforts? 

There are so many opportunities. Having touched on upskilling via TIPs and ed-tech, as well as hiring tools, here are two others:

• Admin tasks – there is a lot of AI anxiety over jobs, for employees and HRs. A common big opportunity is providing tools that reduce admin burdens to focus on value added human-to-human (H2H) activities. For example, AI can draft job descriptions, schedule interviews (which often takes ages), communicate with candidates (which often is hit or miss), help with onboarding, or performance appraisals.
• Talent rediscovery – similar to dating apps, candidate matching is improving using AI. A next step includes matching good candidates who applied for similar roles within a company, but just missed out. They often go into a black hole of lost potential when they would actually fit well for a different new role. AI excels at mass matching.

6. What concerns should companies have when integrating AI into their HCM operations? 

AI is moving at pace and while regulations will be slower, they are coming and will make a difference to HCM. In July 2023 New York regulated the use of AI-driven tools in the hiring process through Local Law 144. This law includes requirements that AEDT (Automated Employment Decision Tools) is ‘bias audited’ and that notices are provided to candidates. While this is a local law, NYC is home to many international companies, and it is likely other jurisdictions will follow in time.

There are other areas of Responsible AI that already apply, or are likely to in time, such as: data protection and privacy; fairness and non-discrimination; accuracy; transparency, explainability; and auditability. Some of these are easy to say but are harder in practice. Many of the LLM (Large Language Models) that recently burst onto the scene are more black than white box solutions, which may be fine for low risk asks, but not for important HR decisions. That said, some providers have been thinking about and integrating Responsible AI solutions for some time.   

7. What can companies do to mitigate bias from AI in HCM?

Bias is an important and interesting topic area. AI systems are trained on data and use data to make decisions. If that data is biased, the systems will be – ‘garbage in, garbage out’. But humans are also biased. We came across an interesting piece of research showing medical school candidates were 10% less likely to be offered a place when they were interviewed (by a human) on rainy days versus sunny days^.[9]

If AI algorithms can be designed to remove subjectivity via training and finetuning, it could be argued AI can help get us to a better place. DEI can improve with better data and measurement. Good implementation, audits and supervision are key. For important decisions, AI recommendations are just that – a tool to help inform, with approvals and decisions made by humans.

8. How can a firm use AI to enhance their HCM? 

AI is not new, but recent developments in deep learning and generative AI has accelerated the capability and usefulness of AI. In turn, there is a great deal of capital and innovation going into the field. This includes HCM solutions. By taking a previously analogue task, such as reviewing a CV, and digitizing it, we create more meta-data to feed AI. This is not just taking place for HR teams but for candidates, so we expect candidates to apply for more roles using AI matching engines (‘Swipe Right’), AI application tools and AI career coaches. This in turn will require HRs use more AI to keep with the volume. The same is true in cyber security, where AI generates more deep fakes and needs AI to find the new attacks. My AI meets your AI.

AI is however more than enhancing a HCM system, it’s about enhancing Human Capital. That AI will substitute some tasks is hardly new news, but we can debate which. A good place to start are the mundane parts and which areas should be upskilled, including for HRs. An aim should be using AI tools to carry out repetitive tasks and using extra time freed up to do more in high-value human areas – HR includes many of these to benefit employees.

An important related area is highlighted by Gallop research that shows positive employee engagements is very poor across firms worldwide (23%).^[10] EX (Employee Experience) drives CX (Customer Experience) which drives revenues and earnings. AI job anxiety is unlikely to help this, but there is an opportunity to re-imagine many knowledge worker roles to both enhance productivity and positive engagement.

9. What should a future-proof HCM system/department need to look like?

Building on the last answer – HCM should help workers thrive.

There are already lots of good innovative AI HCM solutions that can be embraced today, such as TIPs that help move towards skill-centric organizations. Given the pace of change we also have a collective opportunity, with workers, managers and within HR teams, to debate what “good” looks like.

Hybrid work was a test case for this debate. Both AI and remote working are part of the same unstoppable force that is digitization. This force increases speed, personalization, and efficiency. These can all be useful business outcomes, but in terms of future-proofing it is likely that the best talent (workers and HRs) will migrate to the firms that upskill them, reduce mundane tasks and augment, not substitute, them. The best talent will drive business outcomes. You could argue this has always been the case and is more about culture than technology, but the winners are likely to need both – the best of ‘man + machine’ or ‘AI + HRs’