AI in investment advice and robo-advice: overlap with MiFID II
Adopted 2026-06-22 · ≈ 2 min read · 
edited by Dirk Baaijen
Investment advice and robo-advice are not in Annex III of the AI Act, so usually not high-risk. The centre of gravity is MiFID II: suitability assessment, duty of care and transparency. AI does not shift those requirements, but makes explainability and oversight more urgent.
Short answer: AI in investment and robo-advice is not listed in Annex III of the AI Act, so such systems are as a rule not high-risk. The regulatory centre of gravity is MiFID II: the suitability assessment, the duty of care and the information obligations towards the client. AI does not change those obligations, but it makes explainability, data quality and oversight more pressing in practice.
Not in Annex III, so not high-risk (usually)
The AI Act is risk-based. Investment advice and robo-advice do not appear in the list of high-risk uses in Annex III. Unlike credit scoring or life and health underwriting, advice therefore normally falls outside the high-risk regime. The general AI Act provisions still apply, such as transparency about AI interaction and the rules for general-purpose AI systems sitting behind a robo-adviser.
MiFID II is the real framework
Robo-advice remains "investment advice" within the meaning of MiFID II. That brings hard requirements, regardless of whether a human or an algorithm advises:
- Suitability assessment: advice must fit the client's knowledge, experience, financial situation and objectives. An automated questionnaire must elicit and process that information reliably.
- Duty of care and client interest: act in the client's interest, managing conflicts of interest — including in model choice and product selection.
- Information and transparency: the client must understand that and how automated advice is produced, with its costs and risks.
ESMA has previously issued guidelines on automated advice; these remain the benchmark for robo-advice.
Where AI sharpens the requirements
A learning or complex model makes it harder to explain why a given piece of advice was produced, while MiFID II specifically demands accountability. Poor or unrepresentative data lead to unsuitable advice at scale. And a robo-adviser that steers towards in-house products creates a conflict of interest that must be demonstrably managed. This directly touches AI liability where advice causes harm.
What to do
- Classify correctly: confirm the system falls outside Annex III but under MiFID II.
- Secure the suitability assessment: ensure the automated elicitation is complete and up to date.
- Make advice explainable: document how input leads to advice, for client and supervisor alike.
- Manage conflicts of interest: record product selection and steering transparently.
- Connect the governance to your AI governance framework and monitor data quality continuously.
Robo-advice is not high-risk AI, but it is regulated investment advice. The AI Act plays the supporting role here; MiFID II plays the lead.
Sources
- https://eur-lex.europa.eu/eli/dir/2014/65/oj
Directive 2014/65/EU (MiFID II): suitability assessment, duty of care and information requirements for investment advice. - https://eur-lex.europa.eu/eli/reg/2024/1689/oj
Regulation (EU) 2024/1689 (AI Act): risk-based classification; investment advice is not in Annex III.
Read next
Right to explanation of an AI decision: what Article 86 of the AI Act gives you
If you are affected by a decision based (in part) on a high-risk AI system, Article 86 of the AI Act gives you the right to a clear explanation of the AI system's role and the main elements of the decision — from the deployer, on top of your GDPR rights.
AI for strategic workforce planning: usually not high-risk, as long as it does not become individual
AI for strategic workforce planning and skills forecasting at organisation level is usually not high-risk under the AI Act. But once it steers individual decisions, it can tip over. Data quality, governance and transparency remain crucial.
AI, disinformation and elections: deepfakes and the DSA
AI disinformation around elections is caught by several regimes at once: the AI Act's transparency duty for deepfakes, the DSA's risk obligations for large platforms, and existing law. No regime bans disinformation as such.