German Antidiscrimination Chief Calls for Legal Right to Challenge AI Hiring Decisions as EU Locks in High-Risk Rules
04.06.2026 - 08:05:23 | boerse-global.de
A sweeping study of millions of job applications has reignited a debate over automated hiring tools, pushing Germany’s top equal-opportunity official to demand new legal protections. On 2 June 2026, Federal Anti-Discrimination Commissioner Ferda Ataman presented her annual report for 2025 and used the moment to call for an overhaul of the General Equal Treatment Act (AGG). Her central demand: applicants must get a statutory right to know when an algorithm turned them down for a job, a home, or an insurance policy. Current government drafts, she argued, fail to address the discrimination risks posed by automated decision-making.
The sense of urgency is fuelled by fresh evidence of algorithmic bias. Researchers at Stanford’s HAI institute analysed four million job applications from 3.4 million people across 150 employers. They found that one dominant AI screening tool disadvantaged 26 percent of Black applicants and 15 percent of Asian applicants, measured against the so-called “Four-Fifths rule”. The authors warned of an “algorithmic monoculture”: 10 percent of candidates who applied for four different positions were rejected every time – because the same AI tool was used by each employer.
As AI reshapes hiring, the fundamentals of workplace safety and compliance remain a legal bedrock for UK employers. A free toolkit provides ready-to-use risk assessments and checklists to help you meet your duties under the Health & Safety at Work Act 1974 and other key regulations. Download the free Health & Safety at Work Act 1974 Toolkit
Just days before Ataman’s intervention, the European Union moved to cement the regulatory framework around these systems. On 19 May 2026, the European Commission published draft guidelines confirming that AI used in recruitment – from CV screening and candidate ranking to video-based assessments – retains its high-risk classification under the EU AI Act. Internal tools for performance reviews, shift scheduling, and termination decisions are treated the same way. A merely human sign-off, the Commission made clear, is not enough to downgrade the risk level. The classification hinges on the systems’ ability to build profiles.
Those drafts were opened for consultation until 23 June. In a separate move, the Commission closed a consultation on transparency requirements for certain AI systems on 3 June.
Meanwhile, a parallel legislative track is reshaping the compliance timeline. On 2 June 2026, the internal market and justice committees of the European Parliament voted with a clear majority for the “Digital Omnibus” compromise. It introduces staggered deadlines for high-risk AI obligations:
- Stand-alone AI systems: duties apply from 2 December 2027
- Embedded systems in machinery: transition period until 2 August 2028
The aim is to avoid double regulation for products already covered by other EU laws. But some requirements keep the original schedule. Employee training under Article 4 of the AI Act starts on 2 August 2026. The ban on non-consensual intimate images takes effect on 2 December 2026.
Despite the unsettled regulatory landscape, German HR departments are pouring money into AI. SD Worx’s “HR & Payroll Pulse” study, which surveyed 5,900 HR managers and 16,500 employees in 16 countries, found that 48 percent of German HR professionals are currently investing in AI – a sharp jump from 38 percent a year earlier. Mid-sized firms with 250 to 2,499 employees are the most active: 57 percent report targeted AI spending.
Yet there is a gap between technology adoption and oversight. Only 47 percent of German employers have established ethical AI guidelines. For comparison, 50 percent of Dutch HR professionals invest in AI, but just 42 percent have formal ethical guardrails. Across Europe, 37 percent of employers say they would rather deploy an AI solution than hire a new employee – even though 32 percent admit their internal AI knowledge is insufficient.
While HR technology evolves, the core duty to protect employees from physical hazards remains unchanged. A comprehensive Health & Safety Toolkit gives you practical templates for risk assessments, fire safety, and COSHH compliance — all aligned with current UK law. Get the free Health & Safety Toolkit
The market for employee monitoring tools is expanding rapidly. In the first days of June 2024 alone, several vendors launched new platforms: Merge API Inc. unveiled a central hub for AI tools, Workera released an AI agent for skills measurement, and Cybanetix and Adlumin introduced managed-AI security services that can identify “shadow AI” via network analysis. Industry forecasts put the employee-monitoring market at roughly €600 million in 2025, rising to over €1.6 billion by 2034. While 81 percent of companies report productivity gains from tracking, the workforce remains deeply sceptical: 54 percent of employees say they would consider quitting if digital surveillance was significantly expanded. Already, an estimated 71 percent of workers worldwide are subject to some form of digital monitoring.
