AI has moved from an emerging workplace technology to a mainstream business priority in just a few years. Yet, as organisations accelerate adoption, the more important question is no longer whether HR should use AI. It is how technology can improve work without weakening the judgement, trust and human connection on which organisations depend.

This was the central focus of the HONO × People Matters roundtable, “The Future of HR in the AI Era: Balancing Automation and Transformation,” held at The Leela Palace, Bengaluru.

The discussion was opened by challenging a common misconception: automation should not be mistaken for transformation. Removing repetitive work is only the starting point. The real value lies in how organisations reinvest the capacity that technology creates. 

Automation is not the same as transformation

One of the clearest distinctions to emerge from the discussion was between automation and transformation.

Automation creates efficiency. It removes repetitive work, reduces administrative effort and frees up time. Transformation begins with what the organisation chooses to do with that newly created capacity.

This difference matters because it is possible to automate several HR processes without fundamentally improving how talent decisions are made or how the organisation operates. A faster process does not automatically produce better workforce outcomes.

Several transactional activities are already suited to significant automation. AI can support preliminary résumé screening, candidate matching, pre-onboarding documentation, verification workflows, payroll reminders, leave requests and routine employee queries. It can also bring together fragmented workforce data and provide leaders with faster insights into skills, capacity and talent demand.

The real value of these tools is not simply that they reduce manual effort. It is that they allow HR teams to spend more time on decisions that require deeper business and workforce understanding.

A recurring theme throughout the discussion was that complete automation is neither realistic nor desirable. Even in highly digitised functions, there will almost always be a need for human involvement. The right balance between “high-tech” and “high-touch” will depend on the industry, workforce, organisational maturity and consequences of the decision involved.

Human judgement remains critical where trust is involved

The limits of automation become most visible in situations involving emotion, ambiguity or significant consequences for an individual.

AI can shortlist candidates and conduct initial assessments, but hiring does not end with identifying the closest match. Convincing a sought-after candidate to join, understanding their concerns and building confidence in the organisation depend on human rapport.

Career development presents a similar challenge. Technology can recommend roles, skills and learning pathways, but meaningful career conversations require an understanding of an employee’s ambitions, strengths, personal context and uncertainty. These conversations cannot be standardised entirely through an algorithm.

The same applies to employee engagement. Recognition platforms can automate rewards or surface engagement trends, but they cannot replace the role of managers in making employees feel valued. Recognition becomes meaningful when leaders amplify contributions, celebrate achievements publicly and connect individual effort to organisational purpose.

Human judgement is even more important in grievance handling, difficult exits or moments of distress. AI can identify patterns, flag concerns and alert HR teams. It should not independently counsel an employee or attempt to resolve a sensitive workplace conflict.

The dividing line, therefore, is not simply between human tasks and machine tasks. It is between situations that primarily require speed and consistency, and those that demand empathy, discretion and contextual judgement.

AI is redesigning roles and organisational structures

One of the liveliest conversations centred on a question dominating boardrooms today: Is AI replacing jobs, or redesigning them? 

The more immediate shift is in how work is divided, performed and measured. Existing roles are being redesigned as some tasks become automated and employees are expected to use AI to deliver more with greater speed.

This is beginning to reshape traditional organisational structures. The conventional pyramid, built on a large entry-level base supporting fewer experienced professionals, may increasingly give way to a diamond-shaped model. In this structure, smaller groups of highly capable employees use AI to manage wider spans of work and perform tasks that previously required larger teams.

Organisations are responding to this shift differently. Some are reducing entry-level hiring to capture productivity gains from automation. Others continue to hire early-career talent, betting that junior employees equipped with AI tools can take on work that once required far more experience.

Roles themselves are also splitting into two broad categories. AI-augmented roles involve employees using intelligent tools to perform existing work more effectively. AI-native roles are being created around emerging requirements such as AI governance, model oversight, prompt engineering and responsible deployment.

For HR, the challenge is not merely to forecast which job titles may disappear. It is to understand which tasks are changing, which capabilities are becoming more valuable and how employees can move into redesigned roles.

Reskilling must become continuous

As work evolves, reskilling can no longer remain a periodic learning initiative. It must become part of the organisation’s operating model.

The roundtable highlighted the growing anxiety employees feel about job security and relevance. Organisations adopting AI have a responsibility to help their workforces adapt, particularly when technology investments are changing the nature of existing roles.

Employees need more than tool-based training. They need clarity on how their roles are changing, which skills will remain valuable and what new performance expectations will look like. Managers must also learn how to redesign work, guide teams through uncertainty and assess when AI is improving performance versus creating overdependence.

Learning must therefore become more closely tied to business outcomes. As more organisations shift from time-and-material or fixed-price models towards outcome-based contracts, teams are expected to deliver greater value with leaner workforces.

This changes how capability building should be measured. Course completions and learning hours are insufficient. Organisations must examine whether learning improves performance, enables internal mobility, accelerates deployment into priority roles and supports measurable business results.

Governance must be built into AI adoption

The discussion also underscored that responsible AI adoption depends on the strength of an organisation’s data and governance foundations.

HR systems contain highly sensitive information, including compensation, performance, identity and employee records. As privacy regulations become stricter, organisations need clear controls over how this data is used by AI systems.

Participants discussed safeguards such as virtual private clouds, internally hosted models and physically isolated databases. Organisations are also limiting AI tools to verified internal data sources to reduce hallucination and prevent systems from generating unsupported answers.

Yet technical controls alone are not enough. Employees may still upload confidential data into public AI tools for convenience, creating risks that formal policies cannot address on their own.

This makes awareness and behaviour central to AI governance. Organisations need clear usage guidelines, practical employee education and automated blockers that prevent sensitive information from leaving the enterprise environment.

Governance cannot be introduced after deployment. It must be designed into AI-enabled HR systems from the outset.

HR must decide what to do with the capacity AI creates

The larger opportunity for HR lies in how the function uses the capacity created by automation.

As transactional work becomes increasingly digitised, HR can spend more time shaping workforce strategy, organisation design, leadership capability and business performance. But this shift will not happen simply because technology is introduced.

HR leaders must actively evaluate whether AI investments are improving decisions or merely accelerating existing processes. They must assess whether employees are becoming more capable, whether managers are prepared to lead AI-enabled teams and whether workforce structures are being redesigned deliberately rather than only in response to cost pressures.

The future HR function will need to combine technological fluency with human and organisational judgement. It must know where machines can provide speed, scale and intelligence, and where people must continue to provide trust, context and meaning.

Balancing automation and transformation is therefore not a technology decision alone. It is a leadership choice about how work should be designed, how people should be supported and what kind of organisation AI should help create.