Compensation leaders are tasked with paying for roles that continue to change as a result of AI, which means that the job architecture behind them is constantly out of date. The EU Pay Transparency Directive demands that pay be anchored in objective criteria reflecting work as it is actually performed. AI job redesign is moving faster than the job architectures those criteria depend on, which leaves a gap for compensation leaders to close.
I recently attended Gartner’s HR Symposium in London where AI and the workforce were at the centre of the CHRO conversations. Even though the sessions raised plenty of differing views about how to deploy AI technology, there was a common through line: the job, not the tool, is where AI’s value is created. AI releases capacity, but value only appears when the job is redesigned to use it effectively. For global enterprises, the job is also where pay transparency obligations are set.
What Gartner told CHROs about AI and the workforce
Efficiency is the default goal but its returns are limited. Gartner’s opening keynote put numbers on this. Almost 40% of companies are over budget on AI. The organisations getting the strongest results use redesign to change the cost of work, which makes them more than twice as likely to exceed their cost goals, and redirect those gains toward innovation, which raises their odds of outperforming revenue goals by nearly four times. Gartner’s guidance to HR was to concentrate AI investment where it will make the greatest difference.
AI can build jobs people do not want. AI can build jobs people do not want. Gartner’s research on the “bad job” warns that AI is leading employers to design roles that drain people, isolate them or cannot be sustained over time, and that the resulting attrition and talent costs land on HR. Gartner’s advice to CHROs is to spot these roles early and steer the business toward better designs. Each of these roles also has to be paid and compared, which makes them a pay transparency question as well as a retention one.
AI talent carries a premium that may not hold. Where employers pay a premium for AI skills, the value of any one role can shift as the work around it is redesigned. A premium set at the point of offer can outlast the value it was meant to buy. When these AI premiums go ungoverned at the point of offer, they lock organisations into a structural payroll drag that compounds across every subsequent merit cycle, mirroring the broader risks outlined in Syndio’s research on The Hidden Cost of Ungoverned Pay Decisions.
Rehiring is the hidden cost of layoffs. Gartner predicts that by 2029, 30% of employees laid off because AI replaced them will need to be rehired, often at significantly higher cost. The salary premium is only part of the cost. In practice, rehires often return with lower engagement and loyalty, requiring retraining time on top of the institutional knowledge and productivity lost when the original employees left.
How do you align job redesign with objective pay criteria?
Four practices keep job redesign and pay criteria aligned under the EU Pay Transparency Directive.
Integrate pay reviews into the job redesign process. The Directive requires pay structures that can assess the value of the work as it is done now, so pay review has to move with the redesign. Article 4(4) requires pay structures to enable an assessment of whether workers are in a comparable situation in regard to the value of work. The criteria must be objective and gender neutral, and “include skills, effort, responsibility and working conditions.” AI changes two of those directly. Read against Article 4(4), the roles Gartner describes — ones that drain people or isolate them — map to effort and working conditions. A job re-scoped by AI is a different job for evaluation purposes even when the title stays the same. Comparison groups built on last year’s job architecture then misstate who is doing equal work or work of equal value.
Anchor your job architecture in a comprehensive skills catalogue. Article 4(4) names skills alongside effort, responsibility, and working conditions. It adds that “relevant soft skills shall not be undervalued”. Skills such as emotional intelligence, relationship building and ethical judgement are the kind of soft skills the Directive says must not be undervalued.As AI absorbs more routine tasks, those skills increasingly define the job. The comparison still runs between jobs of equal value, so the skills catalogue has to feed the job architecture.
Establish transparent, documented criteria for every skills premium. Article 6(1) requires that employers make easily accessible to workers the criteria used to determine pay, pay levels and pay progression, and those criteria must be objective and gender neutral. A premium for AI skills agreed hire by hire falls short of that standard when the criteria are not written down or visible to workers. The same applies to the premium paid on rehire. Where exceptions are made case by case, gaps open between new hires and incumbents, and the explanation has to be reconstructed after the fact.
Redeploy before reducing. Gartner advises treating the capacity AI frees up as an asset for future growth. For pay, that means redirecting it into higher-value work inside existing roles, then reflecting that in the job and its pay range. Roles redefined this way keep their incumbents, avoid the rehire premium and leave a documented basis for pay.
How should compensation leaders respond to AI job redesign?
Start with the work as it is actually done today, country by country. ADP Research presented at the symposium showed that Europe has the lowest daily AI use among workers of any region (14%) and the lowest share of workers who believe AI will positively affect their job (10%). Adoption is uneven. For a global enterprise, redesign will be uneven too, and the same role may be scoped differently in different countries. A global job architecture needs country-level evidence before it supports pay comparisons.
Compensation leaders can build that evidence with three actions.
- Catalogue the skills the organisation holds today. Start with what exists, by job family and by country.
- Map the catalogue to the current job architecture. Identify where the work has outgrown the job description, then make the plan to close that gap. Do this before running a pay gap analysis, which compares like with like and so compares the wrong people when it runs on outdated job definitions. Once the mapping is done, Syndio can help you determine where pay gaps sit and which to close first.
- Test every skills premium against enterprise value. Where a job requires a premium, document the criteria and the business value it brings before the premium becomes permanent pay.
The work also needs someone to own the criteria for every skills premium, record exceptions consistently, rebuild comparison groups as jobs change, and demonstrate how each pay decision meets the objective, gender-neutral standard.
Pay has to follow the work
The task for employers is to work AI into jobs in a way that is meaningful to the people doing them, without losing their engagement. Pay has to follow that work. When an employee asks why they are paid what they are, the answer should reflect the job they do today, with the criteria documented. This is the foundation for pay transparency that supports compliance and employee trust.
Talk to our team about keeping your job architecture and pay criteria current as roles change.
