EU AI Act: One Hiring Feature Can Make Your AI Startup High-Risk

The same AI model can be low-risk or high-risk under the EU AI Act depending on what you tell customers it does. This short follows Mikael, whose tool summarized job applications. Then a customer asked it to rank candidates. The code stayed the same, but the use case moved it into the employment category of Annex III, which makes it high-risk. - Filtering or evaluating job applicants is listed as high-risk under Annex III (employment) - High-risk means risk management, human oversight, logging and technical documentation - A feature request or landing page claim can change your compliance obligations without any code change Before shipping a new feature, check where your use case lands. You can try LEGALDIGM or book a demo. #EUAIAct #AIAct #highriskAI #AnnexIII #employmentAI #candidateranking #AIcompliance #AIfounders #riskmanagement #humanoversight