Biometric and genetic-data rules are well documented; ADM transparency, AI-risk-assessment, and state-surveillance sub-modules rely on thin or GDPR-baseline-only evidence.
Primary frameworkGDPR Articles 9, 22 as supplemented by the Implementation Act (NN 42/2018)
Supervisory authorityAZOP Traffic-light rationale — AmberBiometric and genetic-data rules are well documented; ADM transparency, AI-risk-assessment, and state-surveillance sub-modules rely on thin or GDPR-baseline-only evidence.
Periodic update · new data 2026-10-25Algorithmic, Biometric & Surveillance Governance
Two developments, one domestic and one at EU level, converge on biometric and surveillance governance in Croatia this reporting period. Domestically, AZOP fined an unnamed gambling operator approximately EUR 2.59 million for unlawfully processing biometric fingerprint data from casino players, a finding that rests specifically on the operator's biometric-identification design: collecting four fingerprints per player without demonstrating the necessity of collecting all four, alongside consent forms that impermissibly bundled multiple processing purposes together. Consent records existed for 34,933 players, underscoring that this was an established, scaled biometric-processing operation rather than a pilot or limited trial.
At EU level, and connected to Croatia through AZOP's own institutional participation, the European Data Protection Board adopted a statement on smart glasses and published a report titled The Connected Gaze at its latest plenary, reported 13 October 2026. AZOP's Deputy Chair, Zdravko Vukić, participated in that plenary. The statement and accompanying report address the surveillance, video-recording and evidence-handling implications that smart-glasses technology raises, a distinct but adjacent biometric-and-surveillance governance concern from the fingerprint-processing case, in that it concerns ambient recording and image capture rather than biometric identifiers collected directly from identified individuals.
The combination of these two developments, a concrete domestic enforcement action against biometric identification practices and direct Croatian institutional participation in an EU-level policy statement on a different but related surveillance technology, signals heightened regulatory attention to biometric and surveillance-adjacent data processing both within Croatia and at the European level during this period. The domestic enforcement action demonstrates that AZOP is prepared to act against established biometric-processing practices that fail to meet necessity and consent-granularity standards, while the EDPB statement demonstrates that the policy conversation around newer biometric-adjacent and surveillance technologies, such as smart glasses, is actively developing at the European level with Croatian regulatory participation.
It is worth noting what is not established by the evidence available: there is no finding connecting the gambling operator's fingerprint-processing practices to the EDPB's smart-glasses statement, and the two developments should be read as parallel signals of regulatory attention to biometric and surveillance technology generally, rather than as connected elements of a single regulatory action.
Outlook
Given AZOP's demonstrated willingness to enforce against biometric-processing design failures domestically, and its direct institutional engagement with the EDPB's developing position on smart-glasses surveillance technology, further Croatian regulatory attention to biometric and surveillance-adjacent processing, whether through additional enforcement action or through domestic guidance responsive to the EDPB's EU-level statement, is a plausible direction for this area going forward.
Periodic update · new data 2026-10-19Algorithmic, Biometric & Surveillance Governance
AZOP's EUR 2.59 million fine against a gambling and casino operator is, at its core, a biometric-governance enforcement action: the operator was found to have collected four fingerprints for biometric identification purposes when only two had been disclosed to data subjects, exceeding what AZOP found necessary for the stated purpose and lacking valid consent. This confirmed decision demonstrates that AZOP is prepared to scrutinise the specific technical scope of biometric data collection against what was actually disclosed to data subjects, rather than assessing biometric processing only at the level of general lawfulness.
Running in parallel with this enforcement action, AZOP began implementing the 2026 European Data Protection Board Coordinated Enforcement Framework (CEF 2026), which this year focuses on GDPR Articles 12 to 14 transparency obligations, with a scope explicitly including AI systems and higher-education-institution transparency practices. A questionnaire response under this framework was due 15 July 2026. This is a confirmed, active coordinated-enforcement exercise that extends AZOP's governance attention into algorithmic and AI-system transparency specifically, beyond the biometric-processing enforcement evidenced by the gambling-operator fine.
Together, these two developments indicate that AZOP's attention to algorithmic and biometric governance is proceeding on two fronts simultaneously: a domestic enforcement action targeting excess biometric data collection, and participation in an EU-coordinated exercise examining transparency specifically in relation to AI systems and the higher-education sector. This dual-track activity suggests a supervisory authority engaging both with individual enforcement matters and with structured cross-border coordination exercises on algorithmic governance within the same period.
Outlook
The outcome of the 2026 CEF exercise on Articles 12 to 14 transparency, once published by the EDPB or by AZOP itself, would be the most informative next development in this area, potentially revealing findings specific to AI-system transparency practices among Croatian controllers. Separately, whether the biometric-collection enforcement approach evidenced in the gambling-operator decision is applied to other sectors using biometric identification remains an open question for subsequent cycles.
Periodic update · new data 2026-10-13Algorithmic, Biometric & Surveillance Governance
AZOP's 2 October 2026 decision against an unnamed gambling operator establishes an applied precedent on biometric data-minimisation necessity testing in a commercial access-control setting. The operator had collected four fingerprints per player, two from each hand, for identification purposes, despite representing to users that it collected only two. AZOP found that the operator could not demonstrate that the larger scope of biometric collection was necessary for its stated identification purpose, and specifically that a justification resting on needing a backup fingerprint, without further substantiation, does not satisfy the necessity test. The fine attached to this finding, within the EUR 2,590,000 total, affected 34,933 players, giving the decision a scale that extends its relevance well beyond the single operator concerned.
The necessity-testing standard applied here is significant for any Croatian controller using biometric identification for commercial access control, identity verification, or similar purposes. It establishes that the stated purpose for biometric collection constrains the scope of what may be collected, and that scope creep beyond what was represented to users, even where framed as a backup or redundancy measure, requires its own independent necessity justification. A controller cannot rely on operational convenience or system-design preference alone to justify collecting more biometric data points than its disclosed purpose requires.
This is a comparatively underdeveloped area of enforcement precedent in Croatia, and the decision gives AZOP's approach to biometric proportionality a concrete, commercially-grounded reference point. Gambling operators specifically, who commonly deploy biometric identification for age verification and self-exclusion enforcement, are a natural audience for this standard, but the necessity-testing logic articulated is not sector-limited on its face.
Outlook
Controllers across sectors using biometric data for identification or access-control purposes in Croatia now have a concrete applied benchmark against which to test their own data-minimisation justifications. Whether AZOP extends this necessity-testing approach to other biometric use cases, such as workplace access control or video surveillance with facial-recognition elements, beyond the gambling-sector context of this decision, is a question this cycle's material does not resolve.
Periodic update · new data 2026-09-28Algorithmic, Biometric & Surveillance Governance
AZOP is reported to have issued guidelines for AI systems involving personal data, emphasising GDPR compliance and risk assessment as the operative framework for controllers deploying AI. The regulator is also reported to have published FAQs addressing legal bases and individual rights specifically in the AI context. Neither the guidelines nor the FAQ document was independently retrieved from a primary AZOP publication this cycle; the finding rests on DataGuidance's aggregator-level summary, which caps confidence at Uncertain rather than Probable or Confirmed.
If the reporting is accurate, this represents Croatia's first documented move toward AI-specific data-protection guidance at the national level, distinct from the general GDPR framework that already applies to any personal-data processing regardless of whether AI is involved. The content of the guidance, specifically what risk-assessment methodology AZOP recommends and how it maps GDPR legal bases onto AI use cases, is not established in the evidence reaching this cycle beyond the general description that it exists and covers these topics.
This is properly read as an emerging rather than settled governance posture. Reports suggest AZOP may have engaged in international coordination on AI and data protection questions, though the specifics of any such coordination were not corroborated by a primary source this cycle and are not asserted here beyond that caveat.
Outlook
The priority for the next reporting period is direct retrieval of AZOP's actual AI guidance document and FAQ text, which would allow this module to move from an aggregator-sourced Uncertain finding to a properly sourced Confirmed or Probable assessment of Croatia's AI-specific data-protection posture.
D8 · AI, Biometrics & Automated — Full Module Record6 sub-modules · 70 words · 11 traced claims · Re-verified 2026-10-25
The decision headline, regulatory posture (Amber), primary framework, and supervisory authority above are open on the Free Explore layer. Register free with a business email for 3 full-record passes a month, or unlock all 170 jurisdictions with Single-Monitor or Portfolio Pro (1 seat) or Firm (up to 8 seats + client-advisory rights).