According to BornCity, the EU AI Office is considering the practical effectiveness of the AI Act and intends to decide soon whether enforcement will be delayed for now. The trigger is the still-pending review of the standards for conformity assessment of high-risk systems. No decision has been made yet.
Background
Since 2 August 2026, operators of high-risk systems are, according to the report, generally required to check their applications for conformity with requirements on transparency, robustness and accuracy, or have them certified. For specific areas, including systems in recruitment or critical infrastructure, further obligations apply only from December 2027 and August 2028 respectively, according to the report.
According to the report, the European Commission published a draft procedure on investigations and fines on 12 October. The consultation ends on 9 November 2026 at 23:59 Brussels time. For prohibited practices, fines of up to 35 million euros or seven percent of worldwide annual turnover are reported, and up to 15 million euros or three percent for other violations; for small and medium-sized enterprises the lower value applies in each case.
Assessment
According to the report, the debate falls into a phase of intensified regulatory activity. On 14 October, EU digital chief Henna Virkkunen asked OpenAI and Anthropic, according to the report, for information on safety measures and model monitoring. Industry bodies such as MedTech Europe call, according to the report, for a unified path that integrates AI Act requirements into existing medical device regulations. If the AI Office loosens the timetable, this could give companies planning certainty; whether and to what extent that happens remains open.
Sources
Editorially owned by Ideal Syka. Sources and method: Newsroom & method. Tips and corrections: ai@i6eal.de.
