Zero-Human Financial Supervision Spain Future Ecb Model
ZERO-HUMAN FINANCIAL SUPERVISION IN SPAIN: THE FUTURE ECB MODEL
Detailed Legal, Regulatory and Technological Explanation with Case Laws
1. Introduction
Zero-Human Financial Supervision refers to a proposed model of financial regulation in which supervisory activities are performed predominantly or entirely through automated digital systems, artificial intelligence (AI), machine learning, real-time data analytics, automated reporting platforms and algorithmic decision-making mechanisms.
Under this model, financial institutions would continuously transmit regulatory, prudential, transactional, liquidity, capital, governance and risk-management information to supervisory systems. Those systems would automatically analyse the information, identify regulatory breaches, detect emerging risks, generate supervisory alerts and, within legally authorised limits, initiate supervisory procedures.
In Spain, the development of such a model would primarily involve the European Central Bank (ECB), the Single Supervisory Mechanism (SSM), the European Banking Authority (EBA), the Banco de España and other competent national authorities.
The expression zero-human supervision should not be understood as permission to eliminate legal accountability or judicial protection. Rather, it describes a highly automated supervisory architecture in which human intervention may become exceptional, while responsibility, transparency, proportionality, procedural fairness and judicial review remain legally protected.
The future ECB model could transform banking supervision from periodic examination into continuous, data-driven and predictive regulatory oversight.
2. Historical Development of Automated Banking Supervision
Traditional banking supervision relied heavily on human inspectors, periodic regulatory reports, documentary examinations, on-site inspections and direct communication between banks and supervisory authorities.
This system had several limitations:
Regulatory information was often received after a reporting delay.
Supervisory authorities depended on manual examination of large quantities of information.
Emerging financial risks could remain undetected until they became serious.
Different supervisory authorities sometimes used different data formats and analytical methodologies.
Human examination could create operational delays and inconsistencies.
The development of digital banking, cloud computing, big data, regulatory technology (RegTech) and supervisory technology (SupTech) created opportunities for automated supervision.
The European regulatory system gradually developed a more integrated supervisory structure through the following developments:
The establishment of the European Banking Authority in 2011.
The introduction of the European System of Financial Supervision.
The creation of the Single Supervisory Mechanism under Regulation (EU) No 1024/2013.
The establishment of the Single Resolution Mechanism under Regulation (EU) No 806/2014.
The implementation of the General Data Protection Regulation.
The introduction of the Digital Operational Resilience Act (DORA).
The adoption of the EU Artificial Intelligence Act.
These developments provide the institutional and technological foundations for more automated supervision.
3. Legal and Institutional Framework in Spain
3.1 European Central Bank and Single Supervisory Mechanism
The ECB is the principal prudential supervisor of significant credit institutions in participating Member States, including Spain.
The Single Supervisory Mechanism operates through cooperation between the ECB and national competent authorities.
Under Regulation (EU) No 1024/2013, the ECB has powers relating to:
Authorisation of credit institutions.
Withdrawal of banking authorisations.
Assessment of qualifying holdings.
Compliance with prudential requirements.
Capital adequacy supervision.
Liquidity supervision.
Governance and internal-control arrangements.
Supervisory reviews and investigations.
Certain enforcement and sanctioning measures.
The Banco de España participates in the SSM and performs supervisory functions under the allocation of responsibilities established by EU law.
A future zero-human supervisory architecture could allow the ECB to receive standardised information from Spanish banks continuously and analyse it through common supervisory algorithms.
3.2 Banco de España
The Banco de España performs national supervisory functions within the European framework and retains responsibilities under Spanish legislation.
Important legal instruments include:
Law 13/1994, of 1 June, on the Autonomy of the Banco de España.
Law 10/2014, of 26 June, on the regulation, supervision and solvency of credit institutions.
Royal Decree 84/2015, implementing Law 10/2014.
Applicable European banking regulations and directly applicable supervisory requirements.
The future model would require the Banco de España to develop secure supervisory data systems capable of interacting with ECB infrastructure.
3.3 European Banking Authority
The EBA contributes to supervisory convergence, technical standards, regulatory guidance and the development of consistent supervisory practices across the European Union.
Automated supervision would require common data definitions, interoperable reporting standards, consistent risk indicators and common approaches to algorithmic governance.
The EBA's role would therefore remain important even if individual supervisory decisions became highly automated.
4. Meaning and Structure of Zero-Human Financial Supervision
A zero-human supervisory system would operate through interconnected technological layers.
4.1 Automated Data Collection
Banks would transmit information through secure and standardised reporting systems.
Relevant information could include:
Capital adequacy ratios.
Liquidity coverage ratios.
Net stable funding ratios.
Non-performing loan data.
Large exposure information.
Credit concentration.
Market risk positions.
Operational risk events.
Cybersecurity incidents.
Governance and internal-control indicators.
Relevant payment and transaction information.
The system would automatically compare submitted information against applicable regulatory requirements.
4.2 Artificial Intelligence-Based Risk Assessment
Machine-learning models could analyse large datasets to identify unusual financial patterns.
For example, an automated supervisory system might detect:
Rapid deterioration in liquidity.
Unusual growth in high-risk lending.
Excessive exposure to particular economic sectors.
Unexplained changes in capital positions.
Unusual related-party transactions.
Increasing dependence on unstable funding.
Possible manipulation of regulatory reporting.
The system could assign risk scores to individual institutions and identify institutions requiring closer examination.
However, an algorithmic risk score should not automatically be treated as proof of a legal violation.
4.3 Continuous Supervisory Monitoring
Traditional supervision often involves periodic reporting and scheduled examinations.
A future automated system could monitor relevant prudential indicators continuously or at frequent intervals.
For example, where a bank's liquidity position deteriorates rapidly, the supervisory system could generate an immediate alert rather than waiting for the next reporting cycle.
This would support earlier intervention and potentially reduce the probability of financial instability.
4.4 Automated Supervisory Alerts
The system could generate alerts when predefined regulatory thresholds or risk indicators are triggered.
Possible alerts include:
Capital adequacy warning.
Liquidity deterioration warning.
Cybersecurity incident warning.
Credit concentration warning.
Unusual transaction warning.
Governance weakness warning.
Reporting inconsistency warning.
Alerts would need to distinguish between technical anomalies, emerging risks and legally established breaches.
4.5 Automated Supervisory Decisions
A more advanced model could automatically process routine supervisory matters, including:
Validation of regulatory returns.
Identification of incomplete disclosures.
Detection of reporting inconsistencies.
Calculation of regulatory indicators.
Verification of predetermined compliance conditions.
Generation of standardised supervisory notifications.
More serious measures, such as withdrawal of authorisation, major capital requirements, administrative penalties or restrictions on banking activities, would require particularly strong legal safeguards.
An automated system cannot acquire powers that the applicable legislation has not conferred on the competent authority.
5. The Future ECB Zero-Human Supervisory Architecture
The future ECB model could operate through five interconnected technological layers.
5.1 Layer One: Unified Supervisory Data Infrastructure
The first layer would create a common European supervisory data infrastructure.
Spanish credit institutions would transmit relevant information through secure reporting channels.
The infrastructure would use common data definitions and standardised reporting formats to reduce inconsistencies between institutions and Member States.
This would help the ECB compare the financial condition of Spanish banks with institutions operating in other participating countries.
5.2 Layer Two: Artificial Intelligence and Predictive Analytics
The second layer would use artificial intelligence to analyse supervisory information.
Potential applications include:
Predictive liquidity analysis.
Early-warning credit models.
Automated capital-risk assessment.
Detection of unusual financial patterns.
Identification of interconnected institutional risks.
Analysis of emerging operational risks.
Detection of possible regulatory reporting manipulation.
Predictive models would support supervisory assessment, but they would not replace statutory prudential requirements.
5.3 Layer Three: Automated Regulatory Compliance
The third layer would compare banking information against applicable legal and regulatory requirements.
For example, a system could calculate whether a bank satisfies applicable capital and liquidity ratios.
It could also identify missing disclosures or inconsistencies between different regulatory submissions.
The system should retain a clear record of:
The applicable legal requirement.
The source information used.
The calculation method.
The algorithmic result.
The supervisory consequence.
Any correction or review performed.
This audit trail would be essential for accountability and judicial scrutiny.
5.4 Layer Four: Automated Supervisory Intervention
The fourth layer could initiate predefined supervisory processes.
For example, a bank might automatically receive a notification identifying a reporting deficiency and requesting correction within a legally permissible period.
However, an automated notification must not be confused with a legally binding administrative decision.
Where an intervention affects an institution's rights or legal position, the system must satisfy the procedural requirements applicable to that decision.
5.5 Layer Five: Legal Accountability and Review
The fifth layer would ensure that automated supervision remains subject to legal control.
It would include:
Traceability of automated decisions.
Access to relevant decision-making information.
Explanation of material adverse outcomes.
Procedures for correcting inaccurate data.
Effective administrative review where required.
Judicial review.
Independent model validation.
Cybersecurity controls.
Clear allocation of institutional responsibility.
This layer is essential because technological efficiency cannot replace the rule of law.
6. Legal Framework Governing the Future Model
6.1 Treaty on the Functioning of the European Union
Articles 127 and 132 TFEU provide important foundations for the ECB's monetary and institutional functions.
Article 127(6) TFEU provides the legal basis for conferring specific supervisory tasks on the ECB concerning policies relating to the prudential supervision of credit institutions and other financial institutions, subject to the Treaty conditions.
The ECB's supervisory powers must remain within the limits established by EU law.
A future automated supervisory model would therefore need to operate within the existing Treaty framework or any lawfully adopted amendments.
6.2 Regulation (EU) No 1024/2013
The SSM Regulation establishes the ECB's supervisory responsibilities.
It provides the institutional framework for direct ECB supervision of significant credit institutions and indirect supervision of less significant institutions through the national competent authorities.
The Regulation permits the ECB to exercise extensive supervisory powers, but it does not create unlimited authority to delegate legally significant decisions to autonomous algorithms.
The future model must distinguish between automated technical operations and the exercise of public supervisory powers.
6.3 General Data Protection Regulation
Regulation (EU) 2016/679 (GDPR) is particularly important where supervisory systems process personal data.
Relevant provisions include:
Article 5: Principles relating to personal data processing.
Article 6: Lawfulness of processing.
Article 22: Automated individual decision-making, including profiling.
Article 25: Data protection by design and by default.
Article 32: Security of processing.
Articles 35 and 36: Data protection impact assessments and prior consultation, where applicable.
Article 22 concerns decisions based solely on automated processing that produce legal effects or similarly significantly affect individuals, subject to its statutory conditions and exceptions.
It should not be assumed that every automated supervisory assessment falls within Article 22. The applicability depends on the nature of the data, the person affected, the legal consequences and the relevant statutory exception.
6.4 EU Artificial Intelligence Act
Regulation (EU) 2024/1689 establishes a risk-based regulatory framework for artificial intelligence.
Certain AI systems used to evaluate the creditworthiness of natural persons or establish their credit score are classified as high-risk systems under Annex III.
This classification does not automatically mean that every AI system used by a banking supervisor is high-risk under the same category.
The legal classification depends on the intended purpose and the relevant provisions of the Act.
Where the high-risk requirements apply, obligations may include:
Risk-management systems.
Data governance.
Technical documentation.
Record-keeping.
Transparency and information.
Human oversight.
Accuracy, robustness and cybersecurity.
Quality-management arrangements.
The Act's implementation timetable and applicable transitional provisions must also be considered.
A zero-human supervisory system would face a particular challenge where the applicable legislation requires effective human oversight.
6.5 Digital Operational Resilience Act
Regulation (EU) 2022/2554, known as DORA, has applied since 17 January 2025.
It establishes requirements relating to digital operational resilience in the financial sector.
Its relevance to automated supervision includes:
ICT risk management.
Incident reporting.
Digital operational resilience testing.
Third-party ICT risk.
Information-sharing arrangements.
Governance and accountability.
A supervisory infrastructure dependent on AI, cloud services and interconnected databases would itself require strong operational resilience.
6.6 Spanish Administrative Law
Law 39/2015 on the Common Administrative Procedure of Public Administrations and Law 40/2015 on the Legal Regime of the Public Sector provide important national administrative-law principles.
Relevant principles include:
Legality.
Competence.
Administrative responsibility.
Procedural guarantees.
Motivation of administrative acts.
Review of administrative decisions.
Electronic administration.
Spanish administrative law also recognises the use of automated administrative action, subject to statutory conditions.
In particular, Article 41 of Law 40/2015 addresses automated administrative action. It requires the relevant administrative authority to identify the competent bodies responsible for defining the specifications, programming, maintenance, supervision and quality control of the automated system, as well as the body responsible for challenging or contesting the relevant action.
This provision is especially important for the future Spanish supervisory model.
It demonstrates that automation does not remove the need to identify the public authority legally responsible for the action.
7. Case Laws Relevant to Zero-Human Financial Supervision
There is currently no established body of judicial decisions recognising a completely autonomous, zero-human ECB banking supervisor.
The following cases provide important principles concerning supervisory authority, institutional delegation, proportionality, automated decisions, transparency and judicial protection.
7.1 Meroni v High Authority (Cases 9/56 and 10/56)
Legal principle: Limits on delegation of public powers
In Meroni, the Court of Justice considered the legal limits on the delegation of powers by EU institutions.
The judgment established important principles concerning the transfer of discretionary powers and the institutional balance under EU law.
Its significance for automated financial supervision is substantial.
An EU supervisory institution cannot avoid legal responsibility by transferring extensive discretionary authority to an external body or technical mechanism in a manner inconsistent with the Treaty framework.
An algorithm cannot independently acquire broad public-law discretion merely because it is technologically capable of exercising it.
For the future ECB model, the distinction between technical assistance and the exercise of legally significant discretionary powers is essential.
7.2 Romano v Institut National d'Assurance Maladie-Invalidité (Case 98/80)
Legal principle: Limits on autonomous rule-making powers
In Romano, the Court examined the limits of powers conferred on an administrative body under the EU institutional framework.
The judgment is relevant to automated supervision because it demonstrates that institutional bodies must exercise powers within the limits established by their legal mandate.
An algorithm cannot create binding legal obligations unless the underlying institutional authority possesses the relevant power under EU law.
The case therefore supports the principle that automated regulatory systems must operate within legally defined supervisory competences.
7.3 United Kingdom v Parliament and Council (Case C-270/12)
Legal principle: Delegation and regulatory authority
This case concerned the legality of powers conferred on the European Securities and Markets Authority (ESMA) under the EU short-selling regulatory framework.
The Court upheld the relevant provisions, finding that the powers were sufficiently circumscribed and subject to legal conditions and review.
The judgment is particularly relevant to automated financial supervision because it demonstrates that EU regulatory authorities may exercise significant intervention powers where the legal framework defines the scope, conditions and safeguards.
A future ECB algorithmic system would similarly require clear statutory boundaries and effective legal controls.
7.4 Landeskreditbank Baden-Württemberg v ECB (Case C-450/17 P)
Legal principle: Allocation of supervisory responsibilities under the SSM
The case concerned the allocation of supervisory responsibilities between the ECB and national competent authorities.
The Court confirmed the structure of the SSM and the ECB's role in ensuring the effective and consistent functioning of banking supervision.
Its importance for Spain is direct.
A future automated supervisory system must respect the allocation of responsibilities between the ECB and the Banco de España.
Automation cannot alter the statutory division of supervisory competence.
For example, an automated system may identify a risk in a less significant Spanish credit institution, but the appropriate supervisory response must follow the legal allocation of responsibilities under the SSM.
7.5 ECB v Crédit Lyonnais (Case C-389/21 P)
Legal principle: Supervisory discretion and proportionality
This case concerned the ECB's assessment of the treatment of certain exposures when calculating the liquidity coverage ratio and the application of the relevant prudential framework.
The Court examined the ECB's supervisory assessment and the legal requirements governing its exercise of discretion.
The judgment is relevant because algorithmic supervision must distinguish between mathematical calculations and supervisory assessments involving legally relevant circumstances.
A purely mechanical algorithm may fail to recognise the individual characteristics of an institution or the specific circumstances relevant to a prudential decision.
The case supports the need for proportionality, reasoned assessment and legally controlled supervisory discretion.
7.6 SCHUFA Holding (Case C-634/21)
Legal principle: Automated decision-making and GDPR Article 22
In SCHUFA Holding, the Court of Justice examined automated credit scoring and the circumstances in which a credit score may constitute an automated individual decision under Article 22 GDPR.
The Court recognised that an automated score can fall within Article 22 where a third party relies on it in a decisive manner when making a decision that significantly affects an individual.
The case is relevant to financial supervision because automated supervisory systems may generate scores concerning borrowers, directors, counterparties or other identifiable persons.
Where those scores are used to produce legally significant decisions affecting individuals, the GDPR requirements must be carefully assessed.
The case does not establish that every automated banking-supervision score is prohibited. Its application depends on the nature and consequences of the processing.
7.7 Dun & Bradstreet Austria (Case C-203/22)
Legal principle: Meaningful information concerning automated decision-making
This case concerned the right of individuals to obtain meaningful information about the logic involved in automated decision-making under the GDPR.
The Court addressed the need to provide information that allows an affected person to understand and challenge an automated decision effectively.
For a future ECB supervisory architecture, this principle is important where automated processing produces decisions affecting natural persons.
The supervisory authority may need to explain the relevant factors and methodology sufficiently to enable meaningful legal challenge, subject to legitimate restrictions protecting confidential information and other legally protected interests.
7.8 Ligue des droits humains (Case C-817/19)
Legal principle: Safeguards for automated data processing
The Court examined the legal framework governing the processing of passenger information by the Passenger Name Record system.
Although the case did not concern banking supervision, it addressed the safeguards required where extensive personal data are processed through automated systems.
The judgment emphasised the importance of necessity, proportionality, objective criteria, independent oversight and protection against unjustified interference with fundamental rights.
These principles are relevant to financial supervisory systems that use large-scale data analytics and automated risk profiling.
7.9 Deutsche Telekom v Commission (Case C-280/08 P)
Legal principle: Judicial review of complex economic assessments
The Court considered the review of complex economic assessments in EU competition law.
Although it is not a banking-supervision case, it is relevant by analogy to the review of algorithmically supported supervisory decisions.
Technical complexity does not remove the obligation to ensure that an administrative decision is based on accurate information, correct legal criteria and a sufficient statement of reasons.
A supervisory authority cannot avoid judicial scrutiny by claiming that a decision was generated by a sophisticated algorithm.
8. Can the ECB Legally Introduce Completely Autonomous Supervisory Decisions?
A distinction must be made between three categories of automated activity.
8.1 Category One: Automated Technical Operations
Examples include:
Data collection.
Regulatory calculations.
Document classification.
Identification of missing information.
Automated reconciliation of reports.
Generation of routine alerts.
These activities are generally more suitable for automation, provided that data security, accuracy and institutional accountability are maintained.
8.2 Category Two: Automated Preparatory Assessments
Examples include:
Risk scoring.
Early-warning indicators.
Predictive liquidity analysis.
Identification of potential capital deficiencies.
Detection of unusual governance patterns.
These activities may support supervisory judgment.
However, their outputs must be appropriately validated and must not be treated as conclusive legal findings without the necessary legal assessment.
8.3 Category Three: Legally Binding Supervisory Decisions
Examples include:
Withdrawal of a banking licence.
Imposition of major capital requirements.
Administrative sanctions.
Restrictions on business activities.
Legally binding corrective measures.
Decisions affecting qualifying holdings.
These measures require particular attention to legal competence, procedural rights, reasoning, proportionality and review.
The ECB cannot lawfully transfer its statutory responsibility to an algorithm simply by describing the resulting decision as automated.
A completely autonomous model for all legally significant supervisory decisions would therefore require a clear and sufficient legal basis and compliance with applicable EU and national fundamental-rights requirements.
Where applicable law requires human oversight or institutional decision-making, those requirements cannot be bypassed.
9. Advantages of Zero-Human Financial Supervision
9.1 Faster Detection of Financial Risks
Continuous monitoring could identify emerging liquidity and capital problems before they become serious.
9.2 Reduction of Manual Errors
Automated data validation could reduce certain errors associated with manual processing.
9.3 Greater Supervisory Consistency
Common algorithms and reporting standards could promote consistency between supervisory assessments across the euro area.
9.4 Improved Cross-Border Supervision
The ECB could compare information from Spanish banks with information from institutions in other participating Member States.
9.5 Predictive Financial Stability
AI systems could identify patterns associated with credit deterioration, liquidity pressure and operational weaknesses.
9.6 Improved Regulatory Efficiency
Automation could reduce repetitive administrative tasks and allow supervisory resources to be directed towards complex cases.
10. Risks and Legal Challenges
10.1 Algorithmic Errors
An algorithm may incorrectly classify a bank as financially unstable because of inaccurate data or unsuitable model assumptions.
Such errors could generate unjustified supervisory alerts or interventions.
10.2 Algorithmic Bias
AI systems may reproduce biases present in historical financial data.
This could create unfair assessments of institutions, customers or particular categories of borrowers.
10.3 Lack of Transparency
Complex machine-learning models may be difficult to explain.
If a bank cannot understand the legal and factual basis of an adverse supervisory decision, its ability to challenge that decision may be weakened.
10.4 Cybersecurity Threats
A fully automated supervisory system would become an attractive target for cyberattacks.
Manipulation of regulatory data could produce incorrect risk assessments across multiple institutions.
10.5 Excessive Dependence on Technology
A supervisory authority that relies excessively on automated systems may become vulnerable to system failures, software defects, cloud-service interruptions and third-party technology providers.
10.6 Concentration of Supervisory Power
A single centralised algorithmic system could create significant institutional risks.
A defect in the system might affect several banks simultaneously.
10.7 Fundamental Rights
Automated systems may process personal data relating to bank customers, directors, employees and other identifiable persons.
The system must therefore comply with applicable data protection, privacy, procedural fairness and judicial protection requirements.
11. Proposed Future Model for Spain
A legally sustainable future model for Spain could be developed through the following stages.
Stage One: Digital Supervisory Infrastructure
The ECB and Banco de España could strengthen secure supervisory data-sharing infrastructure and standardised reporting systems.
Stage Two: AI-Assisted Prudential Monitoring
Machine-learning tools could be used to identify unusual financial patterns and emerging prudential risks.
All material models should undergo validation, testing and independent review.
Stage Three: Continuous Compliance Assessment
Routine regulatory calculations and reporting checks could be automated.
Banks could receive automated notifications concerning identified deficiencies.
Stage Four: Controlled Supervisory Automation
The system could initiate legally authorised routine procedures, while ensuring that decisions involving significant discretion or adverse legal consequences receive the legally required institutional consideration.
Stage Five: Integrated Legal Accountability
The system should maintain complete records of its inputs, outputs, model versions, applicable legal requirements and supervisory consequences.
Affected institutions should have access to appropriate procedures for correcting data and challenging legally binding decisions.
Stage Six: Cross-Border Supervisory Coordination
The ECB could strengthen interoperability between Spanish supervisory data systems and the systems of other participating Member States.
This would improve the detection of cross-border financial risks.
12. Role of Human Oversight in a Nominally Zero-Human System
A fully automated system may be desirable from an operational perspective, but the complete elimination of human accountability would create serious legal and institutional difficulties.
The future ECB model should therefore distinguish between routine automation and legally significant decision-making.
Human involvement may be required for:
Approval of supervisory policies.
Definition of regulatory criteria.
Validation of material algorithms.
Examination of exceptional cases.
Review of contested supervisory assessments.
Decisions requiring discretionary judgment.
Accountability for the exercise of public powers.
The objective should be to minimise unnecessary manual intervention without eliminating legally required safeguards.
This model may be more accurately described as highly autonomous, continuously automated supervision with legally accountable institutional oversight.
13. Institutional Responsibilities Under the Future Model
| Institution | Potential responsibility |
|---|---|
| European Central Bank | Direct prudential supervision of significant institutions and supervisory coordination |
| Banco de España | National supervisory responsibilities and cooperation within the SSM |
| European Banking Authority | Technical standards and supervisory convergence |
| European Commission | Development and enforcement of relevant EU legislative frameworks |
| European Data Protection Authorities | Oversight of applicable personal-data protection requirements |
| European Data Protection Supervisor | Data protection oversight within the EU institutional framework, including the ECB where applicable |
| European Court of Justice and General Court | Judicial interpretation and review of EU supervisory measures |
| Spanish Courts | Review of national administrative and supervisory actions within their jurisdiction |
14. Recommendations for a Legally Sustainable Spanish Model
A future zero-human financial supervisory model should incorporate the following safeguards:
Clear statutory authority: Every automated supervisory action must have a valid legal basis.
Institutional accountability: The ECB and Banco de España must remain responsible for the exercise of their respective statutory powers.
Algorithmic transparency: Material supervisory models should have adequate documentation and explainability.
Independent validation: Critical algorithms should undergo regular technical and legal validation.
Data accuracy: Banks must have procedures for identifying and correcting inaccurate supervisory data.
Proportionality: Automated supervisory interventions must be proportionate to the risk and legally relevant circumstances.
Procedural fairness: Affected institutions must receive the procedural protections required by applicable law.
Effective judicial review: Courts must be able to examine the legal basis and relevant reasoning of supervisory decisions.
Cybersecurity: Supervisory infrastructure must comply with applicable operational resilience and cybersecurity requirements.
Human oversight where required: Automation must not circumvent legal requirements for human supervision or institutional decision-making.
Model governance: The supervisory authority must maintain records of model changes, testing, performance and known limitations.
Third-party accountability: Dependence on external AI, cloud and data providers must not undermine the authority's legal responsibility or operational control.
15. Future Impact on Spanish Banking Regulation
The introduction of highly automated financial supervision could significantly change the relationship between banks and supervisory authorities.
Banks may increasingly operate through continuous digital reporting systems rather than relying primarily on periodic supervisory submissions.
The ECB could identify emerging financial risks earlier, compare institutions more consistently and improve the coordination of cross-border supervision.
However, automated supervision may also increase the importance of data governance, algorithmic accountability, cybersecurity and legal transparency.
Spanish financial institutions would need to invest in reliable reporting infrastructure, high-quality data management, model-risk controls and digital operational resilience.
The supervisory authority would also need to ensure that its technology is reliable, explainable and legally accountable.
The most important legal challenge would be ensuring that technological automation does not weaken the rights of supervised institutions or individuals affected by supervisory decisions.
16. Conclusion
Zero-Human Financial Supervision represents a possible future stage in the digital transformation of European banking regulation.
For Spain, the ECB and the Single Supervisory Mechanism provide the institutional foundation for increasingly automated prudential supervision. The Banco de España would continue to play an important role within the European supervisory structure.
Artificial intelligence, real-time data analytics and automated compliance systems could improve supervisory efficiency, identify financial risks earlier and strengthen cross-border regulatory coordination.
However, the legal framework does not permit supervisory authorities to abandon their statutory responsibilities merely because technology can perform complex assessments.
The principles developed in Meroni, Romano, United Kingdom v Parliament and Council, Landeskreditbank, Crédit Lyonnais, SCHUFA and Dun & Bradstreet demonstrate the importance of legal competence, institutional accountability, proportionality, transparency and effective judicial protection.
The future of Spanish financial supervision should therefore not be based on the uncontrolled replacement of public authorities by algorithms.
Instead, it should develop a system in which routine supervisory functions are highly automated, risk assessments are technologically advanced, and legally significant decisions remain subject to the applicable requirements of institutional responsibility, procedural fairness and judicial review.
The most sustainable future ECB model is not an unaccountable machine-based regulator, but a continuously automated, data-driven and legally accountable supervisory system.
17. Key Legal Authorities and Cases
Legislation
Treaty on the Functioning of the European Union, particularly Articles 127 and 132.
Regulation (EU) No 1024/2013 establishing the Single Supervisory Mechanism.
Regulation (EU) No 468/2014 establishing the framework for cooperation within the SSM.
Regulation (EU) 2016/679, General Data Protection Regulation.
Regulation (EU) 2022/2554, Digital Operational Resilience Act.
Regulation (EU) 2024/1689, Artificial Intelligence Act.
Law 13/1994 on the Autonomy of the Banco de España.
Law 10/2014 on the regulation, supervision and solvency of credit institutions.
Royal Decree 84/2015 implementing Law 10/2014.
Law 39/2015 on the Common Administrative Procedure of Public Administrations.
Law 40/2015 on the Legal Regime of the Public Sector.
Important Judicial Decisions
Meroni v High Authority, Joined Cases 9/56 and 10/56 — limits on delegation of public powers.
Romano v Institut National d'Assurance Maladie-Invalidité, Case 98/80 — limits on institutional rule-making authority.
United Kingdom v Parliament and Council, Case C-270/12 — regulatory powers and institutional safeguards.
Landeskreditbank Baden-Württemberg v ECB, Case C-450/17 P — allocation of supervisory responsibilities under the SSM.
ECB v Crédit Lyonnais, Case C-389/21 P — supervisory discretion and proportionality.
SCHUFA Holding, Case C-634/21 — automated decision-making and GDPR Article 22.
Dun & Bradstreet Austria, Case C-203/22 — meaningful information concerning automated decisions.
Ligue des droits humains, Case C-817/19 — safeguards for large-scale automated data processing.
Deutsche Telekom v Commission, Case C-280/08 P — judicial review of complex economic assessments.
Final legal proposition: Zero-human financial supervision in Spain may develop as an advanced operational model, but it must remain subordinate to EU institutional law, Spanish administrative law, data protection legislation, the AI regulatory framework and the fundamental right to effective judicial protection.

comments