AI in scheduling and shift allocation: what workers and employers should know

Rota software has long helped managers build schedules. Newer tools go further: they forecast demand, decide how many people are needed each hour, offer open shifts to particular workers and, on some apps, hand out each task as it comes in. When software decides who works when, the schedule becomes one of the most direct ways AI shapes people's lives outside work as well as in it. This page covers what that means and what the rules say.

The short answer: AI scheduling can make rotas fairer and easier to change, and can take workers' preferences into account. It can also make hours less predictable, cut breaks and favour some workers over others without anyone noticing. Existing scheduling and working-time laws still apply whoever, or whatever, writes the rota. In the EU, AI that allocates tasks based on individual behaviour or personal traits is listed as high-risk under the AI Act, with duties from 2 December 2027.

What AI scheduling tools do

• Forecasting: predicting footfall, orders or call volumes to decide staffing levels.
• Building the rota: matching shifts to availability, skills, contracts and rest rules.
• Filling gaps: offering open or extra shifts to selected workers, often by app.
• Allocating tasks in real time: deciding which worker takes the next job, route or customer.

EU-OSHA, the EU's workplace safety agency, describes AI-based worker management as systems that gather data on the workplace and workers and feed it into a model that makes automated or semi-automated decisions, including setting routes or schedules (EU-OSHA, AI for worker management, 2022).

Benefits and risks

EU-OSHA lists adjusting tasks or schedules for workers with health needs among the opportunities. Its list of risks includes systems that push workers to skip breaks or work at high speed, and reduced control over how work is done, linked to high stress (EU-OSHA, 2022). For scheduling specifically, the practical risks are:

• Unpredictable hours: a system optimising for demand can change shifts at short notice or cut hours once it sees a quiet day coming.
• Hidden favouritism: if shifts go to the people the system rates as most reliable or fastest, workers with caring duties, health conditions or fewer available hours can lose out without a person deciding that.
• Little room to question it: when nobody can explain why the rota looks the way it does, it is hard to challenge.

See AI and workplace safety for more on pace and stress.

What the rules say

European Union

• High-risk AI: the AI Act lists AI systems intended to be used to allocate tasks based on individual behaviour or personal traits or characteristics, or to monitor and evaluate workers, as high-risk (EU AI Act, Annex III, point 4(b)). Employers deploying such a system must inform workers' representatives and affected workers first (EU AI Act, Article 26(7)). These high-risk duties apply from 2 December 2027, the date set by Regulation (EU) 2026/1744 (European Commission; NI Assembly EU law tracker).
• Predictable work, already in force: under the Transparent and Predictable Working Conditions Directive, where a worker's pattern is entirely or mostly unpredictable, the employer must tell them the reference hours and days within which they may be required to work and the minimum notice before a work assignment. A worker cannot be required to work outside those hours or without that notice, and may refuse such an assignment without adverse consequences. Where an employer cancels an agreed assignment after a set deadline, countries must ensure the worker is entitled to compensation where national law allows cancellation without compensation (Directive (EU) 2019/1152, Articles 4 and 10). Countries had to apply it by 1 August 2022.
• Platform work: the Platform Work Directive treats systems that decide on access to and the organisation of work assignments and on working time as automated decision-making systems, with transparency, human oversight and review rights. EU countries must apply it by 2 December 2026 (Directive (EU) 2024/2831); see EU Platform Work Directive.

United States: an example

Several US states and cities have fair workweek or predictive scheduling laws. Oregon's, for retail, hospitality and food services employers with 500 or more employees worldwide, is one example (Oregon Revised Statutes, ORS 653.412 to 653.485):

• a written good faith estimate of the work schedule at hire;
• the written schedule at least 14 calendar days in advance, and the right to decline shifts added afterwards;
• no scheduling in the first 10 hours after a previous day's shift unless the employee requests or consents, with time and a half for hours worked in that rest period;
• the right to state availability limits, such as child care needs, without retaliation, though the employer does not have to grant the request;
• extra pay for employer-requested changes made without the advance notice, subject to exceptions.

These duties are set by the law, not by how the schedule is made. A tool that ignores them does not move the responsibility off the employer. Local rules vary, so check the law where the work is done. For Great Britain, see software-set rotas and shift changes: the UK rules.

Good practice for employers

• Build the rules in: configure notice periods, rest periods and contract hours as hard limits, not preferences.
• Know what drives allocation: ask the vendor which factors decide who gets shifts, and check outcomes across groups, including part-time staff and people with caring duties.
• Keep a person accountable: a named manager signs off the rota and can change it.
• Tell people: explain to staff, and their representatives, what the system does and how to ask for a change.
• Check forecasts: NIST lists valid and reliable among the characteristics of trustworthy AI (NIST AI RMF 1.0); compare forecasts with what actually happened before cutting hours on them.

See also AI for managers, AI in performance reviews and AI and the right to disconnect.

Questions workers can ask

• Does software decide my shifts or who gets extra hours, and on what basis?
• How much notice will I get of my schedule and of changes?
• How do I record my availability, and who can override the system if it is wrong?
• Who do I talk to if I think shifts are being allocated unfairly?

See AI at work and workers' rights and AI monitoring at work. This page is general information, not legal advice.