As the EU intensifies its focus on digital resilience, data sovereignty, and operational efficiency, backed by regulatory frameworks such as the AI and data acts, organisations across the continent are embracing automation at an increased pace. How can they strike a balance between driving efficiencies while also maintaining effective oversight?
Recent Eurostat data reveals that, as of 2024, 13.5% of EU enterprises with at least ten employees are utilising artificial intelligence technologies, a significant increase from 8% in 2023. This surge is particularly pronounced in countries such as Denmark, Sweden, and Belgium, where adoption rates exceed 24%. The most common applications include text mining, natural language generation and speech recognition, all of which are fundamentally reliant on robust database infrastructure and automated workflows.
The rationale for this rapid adoption is clear. Manual processes, especially those involving data entry, document processing and database management, are not only slow but also fraught with risk.
Research by McKinsey indicates that office workers spend up to half their time handling documents, with around 10% of this devoted to manual data entry. Such inefficiencies are not merely a drain on productivity; they are a source of costly errors. High-profile incidents, such as the accounting misclassification at Macy’s, a US retailer, which resulted in a loss of over $130 million, underscore the dangers inherent in manual data handling.
In the context of the EU, where regulatory scrutiny is intensifying under frameworks such as the General Data Protection Regulation (GDPR), the new Data Act, and the recently enacted AI Act, the cost of manual errors can be even more severe, with fines for non-compliance reaching up to €35 million or 7% of annual turnover.
EU Data and AI Acts and Implications
The EU’s regulatory landscape is evolving rapidly to keep pace with technological change. The EU AI Act, which came into force in August 2024, and the Data Act, set to take effect in September 2025, impose strict requirements on organisations regarding transparency, risk management and data portability. The AI Act not only seeks to monitor the usage of this technology, but also its uptake, with AI literacy now a clear goal. The EU Artificial Intelligence Act, Article 4: AI literacy states: “Providers and deployers of AI systems shall take measures to ensure, to their best extent, a sufficient level of AI literacy of their staff and other persons dealing with the operation and use of AI systems on their behalf, taking into account their technical knowledge, experience, education and training and the context the AI systems are to be used in, and considering the persons or groups of persons on whom the AI systems are to be used.”
Regulations such as the Data Act effectively render manual data management obsolete in most cases for any enterprise that wishes to remain compliant and competitive. Manual database deployments, updates, backups and audits are particularly problematic, onerous and time consuming. They are inherently error-prone and often result in inconsistent environments and failed releases. More critically, they expose organisations to data loss, compliance violations, and cyber threats—risks that are simply unacceptable in today’s regulatory climate.
Given these realities, it is increasingly clear which tasks should never be performed manually anymore. Data entry and document processing, for example, are classic candidates for automation, as they are repetitive, high-volume, and prone to human error. Similarly, database deployments and updates should be automated to ensure consistency, reliability, and rapid delivery of new features or patches. Backups and disaster recovery processes must be automated to guarantee regular, reliable protection against data loss. Security audits and compliance checks, which are essential for meeting the EU’s stringent regulatory requirements, also benefit immensely from automation, enabling continuous monitoring, comprehensive audit trails, and instant reporting. Data governance and quality control, too, are best managed through automated validation and alerting systems, ensuring that data remains accurate, consistent, and compliant with all relevant legislation. According to tech consultancy NCS, even the core processes of extracting, transforming, and loading data—known as ETL—are now routinely automated, enabling real-time, efficient, and accurate data flows across complex organisational systems.
The benefits of such automation are manifold. Automated systems dramatically reduce the risk of human error, accelerate decision-making, and free up valuable human resources for higher-value, strategic work. They also provide the scalability necessary for organisations to grow and adapt in a rapidly changing digital environment. Moreover, automation is now a regulatory imperative. The requirements of the AI Act and Data Act, which demand traceability, transparency, and rapid response to data requests or incidents, are nearly impossible to meet with manual processes alone. Automated systems can generate audit trails, enforce data access policies, and ensure compliance with minimal intervention.
Striking a Balance
However, the question inevitably arises: how much automation is too much? While the benefits are compelling, there are real risks associated with over-automation. One significant concern is the potential loss of human expertise. As more tasks become automated, there is a danger that employees will lose the skills and knowledge necessary to intervene when systems fail or when exceptions arise, or to know how to identify a failure or a problem. This phenomenon, known as “deskilling,” can leave organisations vulnerable, particularly in high-stakes or rapidly evolving situations. There are also ethical and societal implications to consider. Widespread automation has the potential to displace knowledge workers, exacerbating inequality and reducing opportunities for meaningful employment. Furthermore, over-reliance on automated decision-making can lead to what is known as “automation bias,” where errors or biases embedded in algorithms go unnoticed and unchallenged.
The best approach is to strike a careful balance between automation and human oversight. Automation should be deployed to handle repetitive, rule-based, and high-volume tasks—those that are best suited to the speed, accuracy, and scalability of machines. Humans, on the other hand, should retain responsibility for strategic, creative, and high-risk decisions, where judgement, ethical considerations, and innovation are paramount. Collaborative systems, in which humans and AI work together, have been shown to outperform either working alone, combining the strengths of both.
From a regulatory and strategic perspective, compliance is now non-negotiable. The AI Act requires organisations to assess the risk levels of their AI systems, ensure transparency, and maintain human control over high-risk processes. The Data Act mandates data access, portability, and the protection of intellectual property, making manual data management not only inefficient but also legally precarious. Automation, therefore, is not a “set and forget” solution. It requires continuous monitoring, regular testing, and ongoing optimisation to ensure that systems remain effective and compliant. Equally important is the need to train and upskill staff in AI literacy and automation tools, which is both a regulatory requirement and a business necessity.
The landscape for automation and database management in the EU has shifted decisively. Manual handling of repetitive, data-intensive, or compliance-critical tasks is no longer viable. Automation is essential for reducing errors, meeting regulatory requirements, and enabling scalability and innovation. However, there is a clear boundary: automation should never replace ultimate human oversight, expertise or ethical judgement. Instead, the future belongs to organisations that automate the mundane and time consuming while empowering people to lead, innovate, and adapt—ensuring that technology serves as an enabler, not a replacement, for human potential. If a process is repetitive, rule-based, high-volume, or compliance-driven, it should be automated. Manual intervention should be reserved for those areas where human judgement adds irreplaceable value. This is not just best practice; it is now both a regulatory and competitive imperative for organisations.