For many delivery riders, ride-hail drivers and online freelancers, the boss is partly an app. Software offers the jobs, sets the price, tracks the work, rates it and can close the account. The EU's Platform Work Directive, Directive (EU) 2024/2831, is the first EU law written specifically for this kind of algorithmic management. This page summarizes what it says, from the text published in the Official Journal on 11 November 2024 (Directive (EU) 2024/2831).
This is a plain-language summary, not legal advice. The directive sets minimum rules that each EU country must write into its own law, so the details will differ by country.
The short answer: EU countries must apply the directive by 2 December 2026 (Directive (EU) 2024/2831, Article 29). It presumes that a platform worker is an employee when there are facts showing direction and control, and puts the burden on the platform to prove otherwise. It bans platforms from using their systems to process certain personal data, requires them to explain how their systems work, makes them keep people in charge of the decisions that matter, and gives people the right to an explanation and a review.
Who it covers
A digital labour platform, in the directive's terms, is a service provided at least partly online, at the request of a customer, that organises paid work done by individuals and uses automated monitoring or decision-making systems (Directive (EU) 2024/2831, Article 2). The work can be done online or in a physical place, so the definition reaches both delivery and ride-hail apps and online task platforms.
The rules on algorithmic management and personal data protect everyone performing platform work, including people who are genuinely self-employed. Some rights, such as the safety and health duties and information and consultation of workers' representatives, apply to platform workers who are employees or are found to be (Directive (EU) 2024/2831, Article 1).
Employee or self-employed?
Countries must have procedures to establish a person's correct employment status, guided primarily by the facts of how the work is actually done, including the use of automated monitoring or decision-making systems, rather than what the contract calls it (Directive (EU) 2024/2831, Article 4). The relationship is legally presumed to be employment where facts indicating direction and control are found, under national law. The presumption can be rebutted, but it is for the platform to prove that the relationship is not employment (Directive (EU) 2024/2831, Article 5). For how Great Britain, outside the EU, decides the same question, see employment status for app-managed work in the UK.
What platforms may not do with their systems
Through automated monitoring or decision-making systems, a platform may not (Directive (EU) 2024/2831, Article 7):
- process personal data on a person's emotional or psychological state;
- process data about private conversations, including with other platform workers and their representatives;
- collect personal data while the person is not offering or performing platform work;
- process data to predict whether someone will exercise fundamental rights, such as joining a union or collective bargaining;
- infer racial or ethnic origin, migration status, political opinions, religious beliefs, disability, health, trade union membership or sexual orientation;
- use biometric data to identify a person by comparing it against a database of other people's biometric data.
These limits apply from the start of recruitment or selection. Platforms must also carry out a data protection impact assessment of these systems, seek the views of the people affected and their representatives, and give the assessment to workers' representatives (Directive (EU) 2024/2831, Article 8).
Transparency: what people must be told
Platforms must explain in writing, in clear and plain language, which monitoring and decision-making systems they use, what is monitored, the categories of decision the systems take or support, the main parameters and their relative importance, and the grounds for decisions to restrict, suspend or terminate an account or refuse payment (Directive (EU) 2024/2831, Article 9). People must get this in concise form at the latest on their first working day, before changes that affect working conditions, and at any time on request. Workers' representatives get the full detail before the systems are used.
People stay in charge of the big decisions
- Regular evaluation: platforms must oversee their systems and, with workers' representatives involved, evaluate the impact of individual decisions at least every two years, including on working conditions and equal treatment (Directive (EU) 2024/2831, Article 10).
- Staff who can override: the people overseeing the system must have the competence, training and authority to override automated decisions, and are protected from dismissal or other adverse treatment for doing so.
- A human decides on accounts: any decision to restrict, suspend or terminate the contractual relationship or the account, or of equivalent detriment, must be taken by a human being.
- Explanation and review: people can get an explanation of any decision taken or supported by an automated system, and a named contact person to discuss it. Decisions such as suspending an account or refusing payment come with a written statement of reasons (for how this applies to pay set by apps, see algorithmic pay). A person can ask for a review, and the platform must reply in writing within two weeks; if the decision infringed their rights, it must be put right within two weeks or compensated (Directive (EU) 2024/2831, Article 11).
Safety, health and consultation
For platform workers, platforms must evaluate the risks their systems pose to safety and health, including accidents and psychosocial and ergonomic risks, and introduce preventive measures. They may not use these systems in a way that puts undue pressure on workers or puts their physical or mental health at risk, and they must provide effective reporting channels, including for violence and harassment (Directive (EU) 2024/2831, Article 12). Decisions likely to introduce or substantially change these systems are covered by information and consultation of workers' representatives, who may bring an expert; on platforms with more than 250 workers in the country, the platform pays proportionate expert costs (Directive (EU) 2024/2831, Article 13). Where there are no representatives, platforms must inform the workers directly in writing (Directive (EU) 2024/2831, Article 14). See AI and workplace safety for what EU-OSHA has found about AI that sets the pace of work.
How it fits with the EU AI Act
The AI Act separately lists AI used to allocate tasks based on behaviour or personal traits, or to monitor and evaluate workers, as high-risk (EU AI Act, Annex III, point 4(b)). Those high-risk duties apply from 2 December 2027, the date set by Regulation (EU) 2026/1744 (European Commission; NI Assembly EU law tracker). A platform that uses AI to manage workers may have to meet both sets of rules. See AI at work and workers' rights and AI monitoring at work.
Key dates
- 23 October 2024: directive adopted; published in the Official Journal on 11 November 2024.
- 2 December 2026: deadline for EU countries to bring their national laws into force (Directive (EU) 2024/2831, Article 29).
- 2 December 2029: the European Commission must review how it is working, with particular attention to intermediaries and the legal presumption (Directive (EU) 2024/2831, Article 30).
What workers and platforms can do now
- Workers: keep records of how the app assigns, rates and pays your work, and of any account restriction or refused payment. Once your country's law is in force, these are the decisions you can ask to have explained and reviewed.
- Platforms and businesses using them: map every automated system that monitors or decides about people, check it against the data limits, prepare the written information, and decide who can override and review decisions.
- Everyone: watch for your country's implementing law, which sets the final detail. Unions and workers' representatives have a defined role; see AI and unions.
See also will AI replace taxi drivers? and AI policy for employees. This page is general information, not legal advice.
Questions and answers
When does the EU Platform Work Directive apply?
EU countries must bring the national laws that apply it into force by 2 December 2026. The rules reach people through each country's own law, so the exact date and details depend on the country.
Does it make all gig workers employees?
No. It creates a rebuttable legal presumption of employment where facts indicating direction and control are found, under national law. The platform can rebut it, but it must prove the relationship is not employment.
Can an app deactivate a worker automatically?
Under the directive, any decision to restrict, suspend or terminate the account or the contractual relationship, or of equivalent detriment, must be taken by a human being, with a written statement of reasons.
Does it protect self-employed platform workers?
Yes, in part. The rules on algorithmic management and personal data apply to everyone performing platform work, including the self-employed. Some rights, such as the safety and health duties, apply to platform workers who are employees.
How does it relate to the EU AI Act?
They overlap. The AI Act treats AI used to allocate tasks or monitor and evaluate workers as high-risk, with those duties applying from 2 December 2027. A platform using such AI may need to meet both laws.