In most organizations nobody clearly owns AI compliance: the CTO assumes legal has it, legal assumes the CTO does, and HR assumes the vendor does. Fix it by naming one accountable person under one of three models: a dedicated Head of AI Governance, a cross-functional committee with a named chair, or an existing executive with explicit AI responsibility and outside support. The EU AI Act's high-risk obligations apply from December 2, 2027, with fines of up to €15 million or 3% of worldwide turnover.
Free AI leadership role brief
Before you write a job description, get a one-page brief on your role after a 30-minute call: scope, reporting line, a pay range from public filings, and where the candidates are.
Get my role brief →Why does nobody own AI compliance?
The CTO assumes the General Counsel is tracking regulation. The General Counsel assumes the CTO is handling technical compliance. The CHRO assumes AI hiring tools are the vendor's problem. The board assumes management has it covered. This spread of assumed responsibility is the most common and most dangerous AI compliance failure. It persists because the rules are new and still changing, and no traditional role was designed to own them.
What does the EU AI Act actually require?
The EU AI Act has applied in phases since 2025. Its obligations for high-risk systems (AI in employment, credit scoring, education, essential services, law enforcement and critical infrastructure) apply from December 2, 2027 for Annex III uses and August 2, 2028 for Annex I products. What you must do depends on your role.
Providers, who build or sell a high-risk system, need a quality management system, technical documentation (including data, design and performance), conformity assessment, post-market monitoring and serious-incident reporting. Deployers, who use one, must follow its instructions, assign competent human oversight, monitor it and report serious incidents, keep its logs and inform affected workers; some, including those using AI for credit scoring or insurance pricing, must also assess the impact on fundamental rights. Many organizations are both.
What are the penalties?
Maximum fines come in three tiers: up to €35 million or 7% of worldwide annual turnover for prohibited AI practices; up to €15 million or 3% for breaching most other obligations, including the high-risk ones; and up to €7.5 million or 1% for supplying incorrect information to authorities. Someone has to own compliance; if nobody has been named, the organization is exposed.
Why don't existing roles fit?
The General Counsel understands the law but usually lacks the technical fluency to evaluate AI systems or oversee technical documentation. The CTO understands the technology but tends to see compliance as legal's job. The Chief Compliance Officer knows compliance frameworks but may lack the AI knowledge to adapt them. None is wrong for the role, but none is fully right: AI compliance sits across technology, law and risk.
What are the three ownership models?
A dedicated Head of AI Governance
Reporting to the General Counsel, Chief Risk Officer or Chief AI Officer. The clearest accountability and the deepest expertise, at the cost of a new hire. How to find a Head of AI Governance covers the search.
A cross-functional committee with a named chair
Legal, technology, risk and the relevant business units, with a chair accountable for the output and acting as the point of contact. It suits organizations not ready for a dedicated hire but needing structured accountability.
An expanded existing role
Usually the Chief Compliance Officer or a Deputy General Counsel, with explicit AI duties and outside technical support. The cheapest model, but at risk of being under-resourced as demands grow.
What is the first step?
Name a specific, accountable person. Until that happens, the gap grows as the deadlines approach. The EU AI Act: who owns compliance sets out the next steps.
What does a working impact assessment program look like?
Canada’s federal government is a public example. Under its Directive on Automated Decision-Making, departments complete an Algorithmic Impact Assessment before an automated decision system goes live, and the score sets one of four impact levels. By November 14, 2026, 39 assessments had been published: 23 at Level II (moderate impact), 15 at Level I and none at Level III or IV. See all 39, by department and level.
Frequently asked questions
Who is responsible for AI compliance in a company?
One named person should be. Common models are a dedicated Head of AI Governance, a cross-functional committee with an accountable chair, or an existing executive such as the Chief Compliance Officer with explicit AI duties and outside support.
What does the EU AI Act require of companies that use AI?
Deployers of high-risk AI must use it as instructed, assign competent human oversight, monitor it and report serious incidents, keep its logs and inform affected workers. Providers carry heavier duties, including technical documentation, conformity assessment and post-market monitoring.
What are the fines under the EU AI Act?
Up to €35 million or 7% of worldwide annual turnover for prohibited practices, up to €15 million or 3% for most other obligations including the high-risk rules, and up to €7.5 million or 1% for supplying incorrect information to authorities.
Written by Husani Bastien with research and drafting help from AI tools. Every figure was checked against the sources below.
Sources
- Regulation (EU) 2024/1689 (EU AI Act), EUR-Lex: Article 26 (deployer obligations), Article 27 (fundamental rights impact assessment), Article 99 (penalties).
- EU AI Act high-risk dates as amended by the 2026 Digital Omnibus on AI: December 2, 2027 (Annex III) and August 2, 2028 (Annex I).
- Government of Canada Open Government Portal, Algorithmic Impact Assessment records, compiled in Talent Echo, Canada algorithmic impact assessments.