AI’s Potential Role in Employee Discipline Draws Attention in Europe

By Dinah Wisenberg Brin June 15, 2021
AI conceptual image of a face

​Employers' use of artificial intelligence in disciplining workers—or the potential for using it—has started to receive attention in Europe, where various court rulings, proposals and regulations seek to protect employees.

In Spain, a new decree, the Rider Law, which goes into effect in August and amends the country's Workers' Statute, aims to protect food delivery drivers working through digital platforms.

The law establishes that delivery drivers (called "riders") have an employment relationship with companies providing service through online platforms and gives the couriers' representatives the right to information about algorithms or AI systems that affect working conditions and termination decisions, noted Raquel Flórez, an attorney with Freshfields Bruckhaus Deringer in Madrid.

Riders' works councils, or employee representative bodies, will have the right to know the parameters, rules and instructions that the AI systems are based on, she said.

Given how broad the provision is, "the specifics of the information that companies are required to provide are still to be defined," Flórez added. For example, she explained, it remains to be seen whether companies need to provide a report with context on how the AI operates or if the algorithm's raw code is enough.

"Companies are starting to use artificial intelligence in assessing employee performance, but this is still fairly new," said Sonia Cortés, an attorney with Littler in Barcelona, Spain. Digital delivery platforms, however, regularly use AI to manage and assess their delivery riders' performance. "Riders whose performance was below a certain ratio according to algorithms were terminated or were subject to a change in contract terms," she noted.

This led to significant controversy regarding whether these riders were misclassified and should be deemed employees, Cortés said. The Spanish Supreme court ruled in September 2020 that riders were employees.

The Rider Law, enacted in May, applies specifically to drivers delivering products to consumers via digital platform.

European Commission Proposes Rules

More broadly, the European Commission (EC), the European Union's (EU's) executive arm, recently identified AI systems used in employment and worker management as high-risk and proposed harmonized rules that would subject such tools to strict requirements and oversight, Flórez noted. The proposal cites AI systems used in recruitment and hiring; task allocation; and in evaluating, promoting and terminating employees.

Such AI systems should be classified as high-risk because they "may appreciably impact [workers'] future career prospects and livelihoods," according to the EC proposal.

"Throughout … the evaluation, promotion or retention of persons in work-related contractual relationships, such systems may perpetuate historical patterns of discrimination, for example against women, certain age groups, persons with disabilities, or persons of certain racial or ethnic origins or sexual orientation," the EC proposal says. "AI systems used to monitor the performance and behavior of these persons may also impact their rights to data protection and privacy."

GDPR Provisions

While no current EU regulation specifically addresses the use of AI in workplace discipline, the EU's General Data Protection Regulation (GDPR) generally requires companies to seek consent before collecting data on individuals in most circumstances, Flórez noted.

Moreover, the GDPR's Article 22 establishes an individual's right not to be subject to a decision based solely on automated processing, including profiling, that has legal consequences or otherwise significantly affects the person.

Article 22 makes some exceptions, including cases in which an individual has given consent, but clarifications elsewhere in the GDPR suggest a power imbalance in the employer-employee relationship would prevent true consent in labor situations, Flórez said.

The EC doesn't prohibit or limit the use of high-risk AI systems but argues that they should be subject to strict obligations before they can be put on the market and should be subject to oversight afterwards. Plus, they must comply with GDPR Article 22, according to Flórez.

These obligations, she said, include:

  • Adequate risk assessment and mitigation systems.
  • High-quality datasets feeding the system to minimize risks and discriminatory outcomes.
  • Detailed documentation providing all information necessary on the system and its purpose.
  • Clear and adequate information to companies and individuals using the AI system.
  • Appropriate human oversight measures and high levels of security and accuracy.

It remains to be seen how the EC proposal would affect Spain's Rider Law and legislation in other EU countries, Flórez said.

Ruling in the Netherlands

Judges in the Netherlands this year found in favor of app-based ridesharing drivers who complained they'd been dismissed after being subjected to automated decision-making.

The Amsterdam District Court, citing GDPR Article 22, ruled in February that one platform had deactivated drivers based solely on automated processing and ordered the company to pay and reinstate the drivers. The same court in March ordered two ridesharing companies to provide data used to deduct earnings, assign work and suspend drivers.

German Amendment

The use of algorithms in managing platform-based drivers appears to be the exception among employers in Europe, however. Data analytics may help companies make decisions on layoffs but doesn't much come into play for discipline, according to Jan-Ove Becker, an attorney in Littler's Hamburg, Germany office.

Changes in Germany's recently amended works council law include a provision giving councils more involvement in deciding the use of AI in the workplace, Becker said, noting that this provision arose from ongoing European initiatives on developing new AI regulations.

"It is clear that AI is here to stay, and it would not be realistic to think it will not be used as a tool in HR," Flórez said. "AI is not good or bad in itself, but it raises concerns, including, particularly, that it may perpetuate existing patterns of discrimination, among other things. The key will be finding the right balance between finding an efficient way to guarantee EU citizens' fundamental rights while incentivizing innovation."

Dinah Wisenberg Brin is a freelance writer and reporter based in Philadelphia.



Hire the best HR talent or advance your own career.


HR Daily Newsletter

News, trends and analysis, as well as breaking news alerts, to help HR professionals do their jobs better each business day.