On August 2, 2026, the EU’s AI Act entered a new enforcement phase, according to the European Commission’s press release and related guidance. The rules that became applicable cover transparency obligations for AI systems that interact with people or generate or manipulate content, and the Commission’s materials also say recruitment systems are classified as high-risk.
For employers, the workplace rule is more specific: Article 26 of the AI Act says deployers of high-risk AI systems at the workplace must inform workers’ representatives and the affected workers before putting the system into service or using it. The Commission’s service desk also says high-risk deployers must keep logs and monitor operation, while Computerworld reported that the Commission had published guidance for the August 2 transparency deadline.
The immediate pressure point is hiring and other workplace deployments. The supplied sources identify recruitment systems as high-risk and say the transparency rules apply to systems that interact with people or generate or manipulate content. AP also reported that the EU was expanding its AI Office in Brussels by 38 people to monitor AI companies, but that staffing report is separate from the legal duties that took effect on August 2.
What changed on August 2
The European Commission says the AI Act’s majority of rules start applying on 2 August 2026, and the press release says fines can reach up to 7% of global annual turnover for banned AI applications, up to 3% for other obligations, and up to 1.5% for incorrect information. The same press release says recruitment systems are listed as high-risk. Computerworld reported that the Commission published guidelines for the August 2 transparency deadline, and that companies must disclose when users are interacting with AI or exposed to certain AI-generated content.
Those are separate obligations in the supplied record. The transparency rule concerns disclosure to users when they are interacting with AI or encountering certain AI-generated content. The workplace rule concerns notice to workers’ representatives and affected workers before a high-risk AI system is put into service or used at work. The supplied sources do not say these are identical duties, and they should not be treated as interchangeable.
Who is affected
The sources point first to employers and systems used in recruitment and other workplace deployments. The Commission’s materials identify recruitment systems as high-risk, and Article 26 applies to employers using a high-risk AI system at the workplace. Computerworld’s reporting also framed the deadline as a compliance issue for companies using AI in hiring, chatbots, and deepfakes.
Beyond hiring, the supplied sources say the transparency obligations apply to AI systems that interact with people or generate or manipulate content. That scope is broader than recruitment alone, but the research bundle does not identify every product category or every affected sector, so the scope should be read only as the sources state it. The supplied materials also do not identify which specific employers, vendors, or platforms have already changed their practices.
Why the labor angle matters
This is a concrete workplace rule change, not just a policy announcement. The AI Act Service Desk says employers using high-risk AI at the workplace must notify workers’ representatives and affected workers before use. That creates a notice obligation for at least some deployments in workplace settings, including deployments tied to recruitment, where the Commission says recruitment systems are high-risk.
The sources support a narrower claim about labor-market effects than about labor-market outcomes. They show a legal duty to notify workers’ representatives and affected workers, plus transparency duties for certain user-facing AI systems. They do not say the rules will slow hiring everywhere, reduce employment, or change wages. Any such effect would be analysis, not a stated fact in the supplied bundle.
Enforcement and scale
The Commission says the AI Act’s majority of rules start applying on 2 August 2026, and its press release sets out the penalty ranges for different categories of breach. AP reported that the EU was expanding its AI Office in Brussels by 38 people to monitor AI companies. The supplied sources do not say how those 38 people will be assigned or what cases they will handle.
There is also uncertainty in what the sources do not identify. They do not name the full set of employers or vendors now subject to the transparency duties, and they do not specify how many companies already have compliant notice systems in place. The record also does not identify which member-state authorities, if any, will handle the most immediate workplace disputes under Article 26.
What to watch next
The next questions are practical rather than abstract. Employers using high-risk AI at work will need to know when notice to workers’ representatives and affected workers is required before use, and companies offering AI that interacts with users will need to disclose that interaction or exposure where the transparency rules apply. The supplied sources also point to the Commission’s guidance as a reference point for deployers facing the August 2 deadline.
The record does not say how quickly compliance will spread, which sectors will adapt first, or whether companies will redesign hiring tools to avoid the high-risk category. It also does not say whether the AI Office’s staffing increase will lead to faster audits or more enforcement actions in the near term. Those remain open questions based on the supplied reporting.
Source limits and scope
All of the claims above are limited to the supplied bundle. The European Commission press release supports the start date for the majority of rules, the high-risk status of recruitment systems, and the penalty ranges. The AI Act Service Desk supports the workplace notice duty and the log-and-monitor obligations for high-risk deployers. Computerworld supports the existence of Commission guidance for the transparency deadline and the disclosure duty for AI interactions or certain generated content. AP supports the report of a 38-person expansion of the AI Office in Brussels.
Where the sources do not identify something, that absence is preserved here as absence. The supplied sources do not identify every affected employer, every affected AI product, or every enforcement timetable beyond the dates and obligations stated above.
Sources
- European Artificial Intelligence Act comes into force — European Commission
- Article 26: Obligations of deployers of high-risk AI systems — European Commission AI Act Service Desk
- The EU's AI transparency deadline is weeks away. Is your enterprise ready? — Computerworld
- EU begins crackdown on AI risks — AP News