At a time when artificial intelligence is deeply permeating policy, operations and public services, data quality is the backbone of reliable and legitimate digital decision-making. AI systems learn, reason and predict based on data; but when that data is incomplete, outdated or biased, not only do outcomes come into question, but also the legitimacy and ethical integrity of the technology. Errors in datasets can lead to discriminatory algorithms, incorrect analysis, enforcement risks under the AVG and AI regulation, and significant reputational damage.

This dossier 'Data Quality & Compliance' focuses on the question of how organizations can structurally safeguard data quality in an era of tightened AI regulation, digitization and increasing public and supervisory control. It is emphatically about the interaction between legal standards, technical choices and organizational governance around data and algorithms.

Topics covered include:

• The core principles of data quality: accuracy, completeness, timeliness, consistency and traceability of data within complex data streams.
• The relationship between data quality, bias, explainability and reliability of AI systems in both public and private domains.
• Governance, validation, data cleaning and auditing as tools to ensure data quality throughout the AI lifecycle - from design and training to deployment and monitoring.
• The legal frameworks and responsibilities under the AVG, the AI Regulation, now in force, and relevant industry standards and guidelines around data responsibility and algorithm oversight.
• The role of data controllers, data stewards, CDOs and compliance officers in establishing oversight of data flows, risk management and reporting to the board and regulators.
• Practical lessons from incidents in which organizations faced legal, social or ethical consequences of poor data quality.

In addition, the dossier focuses on current issues such as the use of synthetic and generated data, the validation and documentation of training data, the integrity of data when deploying foundation and GPAI models, and the growing reliance on external data providers and data ecosystems. This brings into sharp focus how technology, governance and trust are inextricably linked in a mature data quality and compliance strategy.