10 min read
Artificial Intelligence’s impact on work depends on how organisations translate its capabilities into operational decisions. Expected time savings should be assessed alongside the work that remains, the skills employees need and the consequences for workload and staffing.
AI can reduce the time employees spend on particular tasks and the business value depends on what happens next. Employers need to decide how that time will be used, what work remains and how responsibilities, workload and employment will change.
These questions are relevant to organisations of different sizes, including those without a (European) Works Council. It provides a basis for discussion between HR, management, operational teams and employees about the choices behind AI implementation.
In this article we look into how expected efficiencies with AI translate into daily operations and what they mean for employers and HR.
AT A GLANCE
What Is the EU AI Act?
The AI Act Regulation (EU) 2024/1689 aims to support the use of Artificial Intelligence while protecting health, safety and fundamental rights. It regulates AI systems and establishes separate obligations for providers of general purpose AI models. The AI Act is an EU regulation and it applies directly in EU Member States.
The Act applies to all EU Members states and organisations that develop, supply or use AI systems within its scope. This includes:
- Providers: organisations that develop an AI system, or have one developed, and make it available under their own name. For example, a software company offering an AI recruitment tool to employers in the EU.
- Deployers: organisations that use an AI system in their professional activities. For example, an employer using that recruitment tool to assess job applications.
- Importers and distributors: organisations that bring AI systems from outside the EU onto the EU market or supply them within the EU.
The Role of Employer and HR in AI
- The employer and management set the business objectives, approve investment, decide how released capacity will be used and remain accountable for organisational decisions.
- HR assesses the implications for roles, staffing, workload and skills, advises management, and coordinates training, employee involvement and relevant HR processes.
- Together, they evaluate results and adjust the approach when the expected benefits or employee outcomes are not achieved.
Connect AI Investment to Business Objective
A potential business investment in AI should address a specific business need, such as reducing response times, improving service quality or managing a growing workload. Before choosing a tool, management should identify what causes the problem and assess how AI could help. Faster drafting, for example, may have little effect on response times if work continues to wait for approval.
HR may support management assess how the proposed investment would change employees’ tasks, responsibilities and workload. This includes identifying the skills, training and support needed to achieve the intended improvement. Any expected benefits should account for the time employees will spend checking outputs and correcting errors.
The aim is to establish how AI will improve the way work is performed and what the organisation needs to make that improvement possible.
Discussion question: What business problem should AI help us solve, and how will we measure its effect on performance, service quality and employee workload?
Assessing Time Savings and Staffing Decisions
Time saved by an AI tool on a task is not automatically time that can be removed from a job position. Small savings of time spread across a working day may be difficult to combine into a usable block. It means that employees may still need to remain available for customers, peak demand or specialist work.
For example, AI may produce a first draft quickly while creating additional checking, correction or approval work elsewhere. Any estimate should account for these activities, together with training, system costs and ongoing support.
Another example, saving ten minutes on drafting does not release ten minutes if an additional six minutes are needed to verify the output of AI. The remaining four minutes may therefore not translate into a staffing reduction if one of the tasks in the position still must provide continuous cover.
Discussion question: How much time is saved across the complete end-to-end process, and can that time realistically be used for something else?
Organisational Choices Shape AI Outcomes
TNO research published in June 2026 and reviewed literature, found that productivity and employee outcomes varied by application and by decisions about task allocation and work organisation. The study offers useful examples, without establishing that every employer will achieve the same results.
Released capacity of employees and time due to the implementation of AI may support the organisation in:
- growth.
- clear backlogs.
- improve customer contact.
- reduce reliance on overtime and external support.
However, it may also lead to role redesign, retraining or staffing changes and potential redundancies.
The practical implication for employers is to discuss the intended use of time explicitly. If an AI project assumes both lower staffing costs and more time for development, management should explain how the available capacity supports both objectives.
Discussion question: Are we using the expected capacity to improve output, reduce costs, improve working conditions or pursue a combination that we can substantiate?
Workload and Autonomy
Removing (routine) work and automation with AI, can make a job role more interesting. However, it can also leave employees handling difficult cases throughout a larger part of the day. Customer complaints, exceptions and complex decisions may require more concentration than the tasks AI has taken over.
HR and operational managers should therefore assess the composition of the remaining work. Work volume alone may not reveal changes in mental demands, recovery time, autonomy or access to support for employees.
The employee experience as well as performance expectations are useful evidence here. Employees may identify checks, workarounds and exceptions that are absent from a process map or a management forecast.
Discussion question: What will a normal working day look like after implementation, and will employees have the time and support to perform the remaining work well?
Autonomy and Skills
Employees who review AI outputs need the skills to recognise errors and the authority to act on their findings. They should understand the system’s limitations, know when to question an output and have access to specialist support. Approval by an employee or manager is meaningful only when employees have sufficient information and time to review the result and can correct, reject or escalate it.
Training offered to employees should reflect how employees use AI and the decisions they are expected to make. An introductory session may cover basic use, while employees responsible for reviewing outputs or supervising others may need further training to assess reliability and resolve concerns.
Employers should also consider how AI changes learning within a role. How will less experienced employees develop their judgement and progress in their roles if AI takes over tasks that previously provided opportunities for learning and supervised practice?
Discussion question: Who is accountable for the outcome, and do the employees involved have the skills, information, time and authority to fulfil their responsibilities?
Data Use Privacy and Employee Confidence
Employees need to understand what information an AI system uses, why it is used and how its outputs may affect their work. This is particularly important when AI supports performance assessment, task allocation or decisions about employees.
HR should work with the IT team to establish clear rules for collecting, accessing and using this information. Where personal data is processed, the GDPR requires a lawful basis, a defined purpose and appropriate safeguards. Employees must receive the required information about the processing. A data protection impact assessment is required where the processing is likely to pose a high risk to individuals’ rights and freedoms.
The EU Data Act may also be relevant where an organisation accesses data from connected equipment, devices or related services. However, access to that data does not automatically permit its use to monitor or assess employees. Any processing of personal data must still comply with the GDPR and actions from the employer on AI are required.
In practice, employees need clear instructions about approved AI tools, the information they may enter and how to report inaccurate or inappropriate outputs. Management should assign responsibility for investigating concerns and deciding when the system’s use needs to be changed or paused. These arrangements should remain effective after implementation, as the system’s use and its effects on employees develop.
Discussion question: Do employees understand how their information is used, how AI affects decisions about their work and where they can raise a concern?
Evidence to Support the Decision
When management considers changes in the organisation, based on expected AI efficiencies, it should explain what supports those expectations. A forecast, a pilot and established operational results provide different kinds of evidence.
Forecasts
- A forecast estimates what management expects to achieve. It may draw on supplier information, assumptions about employee adoption or calculations of potential time savings.
Pilots
- A pilot shows what happened within a defined trial. Its findings should be considered alongside the conditions under which they were achieved. Participants may have received additional support, volunteered because they were confident using AI or worked with a limited selection of tasks. These conditions can affect how far the results can be applied to the wider workforce.
Operational Results
- Operational results show how AI performs in everyday use. Relevant measures may include time spent completing the whole process, output quality, corrections, customer outcomes and employee workload. Faster completion of one task can be useful, but its value depends on the effect on the complete process.
Relevant measures include the time required to complete the whole process, output quality, corrections, customer outcomes, workload and costs. A rise in tool usage demonstrates adoption; it does not by itself demonstrate a potential business improvement.
Where staffing changes are considered, distinguish the contribution expected from AI from other reasons such as lower demand, duplication or a change in service strategy. This supports management assess what a potential business proposal regarding AI depends on and what happens if the expected efficiencies are delayed.
The same AI tool may be used in different languages, roles and customer environments. Teams may also differ in data quality, existing digital systems, work volumes and experience. A result from one location should therefore be assessed before it is used as a planning assumption elsewhere.
Discussion question: Which expected benefits have been demonstrated in everyday work, and which are still forecasts?
Connect AI Decisions to Works Council Consultation
An organisation with a European Works Council or local Works Council must follow Works Council consultation before implementing any organisational changes or AI implementation tools. Any company without an EWC or local council may also still have obligations towards national employee representatives or individual employees.
Management should obtain legal and HR advice on any required information, consultation or consent procedures before committing to implementation. HR can coordinate this assessment with legal, privacy and technology specialists and ensure that the resulting requirements are reflected in employee communications, training and working practices.
AI Agenda Future of Work
AI is changing how international organisations allocate tasks, organise work and plan their staffing needs across Europe. Employers, HR and management and Works Council will need to keep AI and the Future of Work on to the agenda. These discussions can gain clarity on how expected company efficiencies from AI may translate into daily operations and what they mean for employees locally and across borders.



