How the AI Act lands in specific industries: which Annex III or Annex I categories catch common uses in HR, lending, insurance, healthcare, education, critical infrastructure, the public sector, SaaS and regulated products, and the sector-specific carve-outs and dates.
Why recruiting, promotion, task allocation and worker monitoring AI is high-risk, the ban on emotion recognition at work, what employers as deployers must do (worker information, oversight, logs, explanations) and what HR software vendors carry as providers.
AI Act duties for creditworthiness and credit scoring systems: high-risk status and the fraud detection carve-out, fundamental rights impact assessments, financial-sector QMS, documentation and log rules, financial supervisors as market surveillance authorities.
How the AI Act treats insurance: high-risk status for life and health risk assessment and pricing only, the fundamental rights impact assessment for those deployers, financial institution carve-outs, explanations to customers and the 2027 date.
AI Act rules for health: AI in medical and in vitro diagnostic devices through Annex I and the August 2028 date, the sectoral conformity route and single technical file, emergency triage and benefit eligibility in Annex III, and incident reporting under device law.
Which education uses are high-risk (admission, grading, level assessment, exam proctoring), the ban on emotion recognition in schools, what institutions as deployers and edtech vendors as providers must do, and the 2027 date.
AI Act duties for safety components in digital infrastructure, road traffic, water, gas, heating and electricity: high-risk status, internal control, national rather than EU database registration, no fundamental rights impact assessment, testing and supervision rules.
What public authorities using AI must do: fundamental rights impact assessments, registering their use in the EU database, benefit eligibility and emergency systems, the social scoring ban, explanations to citizens and the 2030 deadline for legacy systems.
How a SaaS vendor that adds AI features is treated: provider or deployer, chatbot disclosure and content marking, when a feature becomes high-risk or exempt, contracts with model suppliers and becoming the provider of a repurposed system.
AI Act duties for product manufacturers: AI as a safety component under Annex I, the 2026 non-safety clarifications, machinery moving to Section B, the August 2028 date, the sectoral conformity route, manufacturers as providers and real-world testing frameworks.
Tell us what you're working on, whether it's a deployment, an audit, a security test or a cyber range. You'll speak with an engineer who can help you scope it.