A new analysis of workforce strategy at seven global employers argues that artificial intelligence is reshaping jobs at the task level far more than it is eliminating them outright, and the functions absorbing the redesign work are the ones inside HR itself.
The reallocation thesis
TalentNeuron, a workforce intelligence firm, released a report titled “The Great Reallocation: Understanding the Impact of AI on Talent Strategy” this week, built from an analysis of 114,419 global job postings requiring core AI skills across 103 occupations, combined with case studies inside seven enterprises: Salesforce, Klarna, Wells Fargo, Google, Microsoft, Citi, and BT Group. The research also draws on labor market work from McKinsey and Company, the OECD, MIT’s NANDA initiative, and the Anthropic Economic Index.
The headline finding cuts against the layoff narrative that has dominated AI and jobs coverage for two years: no job in the dataset was fully automatable. At a Fortune 100 manufacturer TalentNeuron studied in detail, 34% of roles that had initially been flagged for elimination turned out, once broken down task by task, to contain work that required human judgment critical to the company’s own transformation strategy. The roles were not safe as originally designed, but they were not disposable either. They needed to be redesigned around the judgment calls a model could not make.
Where the demand is actually rising
Inside the seven companies TalentNeuron studied, demand for HR, strategic workforce planning, and people analytics roles rose a combined 16% over two years, driven by three specific categories: strategic workforce planning skills, up 33%; people analytics, up 26%; and learning and development specialists, up 42%, the steepest increase of the three. Those are not entry level postings. They are the roles a company needs staffed correctly before it can credibly redesign anyone else’s job, which is part of why demand for them is compounding rather than flattening.
What that means for the shape of HR itself
“AI is accelerating the pace of workforce decisions, compressing planning cycles from years to quarters, and sometimes even weeks,” said David Green, co-founder and managing partner at Insight222, in the report. Workforce planning built for an annual cycle cannot answer a question that changes shape every quarter, which is part of the practical reason the report’s data shows planning and analytics headcount growing even as adjacent operational roles get redesigned or absorbed.
Erzsebet Malzenicky, global head of workforce strategy and transformation at Experian, put the underlying problem in more granular terms: “You can’t design a workforce that blends human and automated capability without knowing, at a granular level, what your people actually do.” That is a task mapping problem, not a headcount problem, and it explains why the report finds people analytics and workforce planning functions growing in the same window that other roles are being restructured around automation.
The task level test
The report’s central methodological point is that job titles are the wrong unit of analysis for AI impact. A role survives or gets redesigned based on what fraction of its actual tasks require judgment, context, or accountability a model cannot assume, not based on what the job is called. “The organizations that will benefit most are those that understand how work is performed at the task level,” said David Wilkins, chief executive of TalentNeuron. Companies that skip that task level audit and instead cut by job title, the report implies, are the ones most likely to eliminate work they will have to rehire for within a year.
What it means for the HR leader
Three things follow directly from this data for anyone running workforce planning right now. First, a redesign audit has to happen at the task level before headcount decisions are made, not after; skipping it is how a company ends up cutting the 34% of a role that actually mattered. Second, the internal functions doing that audit work, workforce planning, people analytics, and learning, need to be staffed ahead of the redesign wave, not scaled in response to it, because the seven company data shows demand for them rising before the broader restructuring plays out. Third, the planning cycle itself needs to shrink; an annual workforce plan cannot track a quarter by quarter reallocation, which is the compression Green describes above.
There is a harder implication too. If 34% of roles flagged for elimination at one manufacturer actually contained judgment work the company could not do without, then a large share of the layoffs already announced across the industry this year were probably measuring the wrong thing. Companies that cut by job title rather than by task are not just risking morale, they are risking having to rebuild capability they just paid severance to remove.
What to do next
HR and workforce planning leaders evaluating their own AI rollout should start with a task level inventory of the roles under the most transformation pressure before committing to headcount targets, and should treat the workforce planning and people analytics functions themselves as the first roles to properly resource, not the last. Related coverage on this site has tracked the operational side of the same shift: internal misalignment slows AI rollouts more than the technology itself, and the wage premium for AI adjacent skills keeps climbing even as automation anxiety dominates the headlines.
Managers Are the Weak Link in AI Rollouts and The AI Skills Wage Premium Just Hit 62% cover related ground on how AI adoption is playing out inside HR functions.
Source: TalentNeuron
