After more than two years of legislative highs and lows for the European Union’s proposed Platform Work Directive, the EU may be inching closer to adopting regulations on the classification of platform workers, such as rideshare drivers and app-based delivery persons, as Belgium assumes the presidency of the Council of the EU.
If adopted in its current form, the Directive would (1) create a rebuttable presumption that platform workers are employees when two of five indicators are present, and (2) require human oversight and transparency when using artificial intelligence (AI) to make decisions vis-à-vis those platform workers, such as allocating work or terminating accounts.
Quick Hits
- The European Parliament and Council are expected to soon reach an agreement on the final text of the EU Platform Work Directive, which, if adopted as currently drafted, will create a rebuttable presumption that platform workers are employees, not independent contractors.
- The Directive would also require platforms to notify platform workers when using AI in a way that could impact them, and incorporate human oversight into the process.
The Directive Is Not Yet Law
Although EU legislators reached a provisional deal on December 13, 2023, it was quickly followed by the Spanish presidency’s December 22, 2023, conclusion that majority approval of the agreement could not be reached. Rather than put the issue to a formal vote, the Spanish presidency deferred further action on the matter for the Belgian presidency, which began on January 1, 2024. Discussions continued on January 16, 2024, and while a decision has not yet been reached, an agreement as to the Directive’s scope and terms is anticipated, given the increasing number of labor courts throughout Europe that have determined that platform workers are misclassified employees.
Cracking Down on Misclassification
As drafted, the Directive would create a rebuttable presumption of employment (rather than self-employment) if at least two out of the five following indicators of direction and control are present:
- “upper limits on the amount of money workers can receive”
- “supervision of their performance, including by electronic means”
- “control over the distribution or allocation of tasks”
- “control over working conditions and restrictions on choosing working hours”
- “restrictions on their freedom to organize their work and rules on their appearance or conduct”
Member states would also have the ability to expand on the list of indicators as they see fit, which would create more opportunities to claim the presumption. In fact, the original draft required a presumption of employment if three out of seven indicators were present, and some member states already have similar legislation on platform work with additional indicators, so this list may change if and when an agreement is reached. Any such additional indicators are likely to follow the same theme of assessing the level of “direction and control” a service recipient retains over the service provider, which is consistent among tests for identifying misclassification globally.
Further, under the current draft, each member state’s authorities would have the ability to initiate a misclassification claim on its own, and to decide if it will apply in tax, criminal, and social security settings, as well as in administrative and judicial proceedings. To give bite to that bark, the directive (in its current form) requires platforms to notify national authorities regarding who works for them and what types of activity they do. It is not yet clear what the consequences might be for failing to comply with the notification requirement.
The text of the directive is still in flux, so this might not be reflected in the final rule.
AI Transparency and Human Oversight
When introduced in 2021, the directive was positioned to provide the first-ever EU rules on AI in the workplace, but the EU AI Act (expected to become law soon) beat it to the punch when a deal was reached on December 8, 2023. Nevertheless, the Directive is in line with the EU AI Act’s determination that AI in the workplace is a “high risk,” and it outlines requirements for users of AI, rather than creators of AI, which is where the EU AI Act falls short. To that end, the Directive has the potential to serve as a guidepost as to the expectations of legislators when it comes to employers using AI.
As currently drafted, the directive would:
- require each platform to provide platform workers, whether employees or contractors, with a written explanation of how AI is used within the platform to: (a) monitor, supervise, or evaluate work performance, or (b) make decisions affecting the platform workers’ working conditions, such as access to work assignments, earnings, health and safety, working time, promotion, and contractual status, including restriction, suspension, or termination of their accounts;
- prohibit platforms from processing personal data of platform workers that is not “intrinsically connected to and strictly necessary for” the performance of the work; and
- require platforms to have human oversight of AI, including providing platform workers that have been impacted by “decisions taken or supported by” AI with explanations for such decisions.
To ensure compliance with the AI transparency and human oversight regulations, the Directive (as currently drafted), contemplates fines of up to €20,000,000, or four percent of total annual turnover in the preceding financial year, whichever is higher.
Next Steps
The text of the directive is still in flux, so the provisions of the current draft may not reflect the final rule.
Employers using AI tools, and particularly those with cross-border operations, may wish to follow these developments and evaluate proactive compliance efforts.