Malta's AI ITA Certification Scheme — A Voluntary National AI Assurance Framework With Years of Near-Zero Uptake

In October 2019 the Malta Digital Innovation Authority (MDIA) launched the AI Innovative Technology Arrangement (AI ITA) scheme, describing it as the world's first national AI certification programme. Under the scheme, developers and deployers could have AI systems certified by an MDIA-licensed 'systems auditor' as meeting pre-set ethical, transparency and technical criteria — a voluntary conformity-assessment model that anticipated elements later formalised in the EU AI Act.

To provide a voluntary national AI certification and assurance pathway ahead of binding EU-level regulation.

MDIA subsequently issued Technology Assurance Sandbox programme guidelines (2021) and blueprint guidelines (2022), and opened a public consultation on an ITAS Sandbox in 2024. Subsidiary Legislation 591.01 (the ITAS Certification Regulations) entered into force on 14 June 2024, making the certification framework 'fully operational' under Malta's regulatory tracker. MDIA is now developing a further Technology Assurance Assessment Framework, with consultations planned through 2025.

An independent academic assessment of AI certification schemes found that, as of 2021, no application had yet been submitted under the scheme — two years after launch, the flagship programme had attracted no real-world adopters. As of the most recent public documentation, no statistics on the number of certifications issued, sandbox participants, or applications under the revised regime have been published.

This is included as a cautionary case: a well-publicised, carefully designed voluntary AI-assurance regime that, five years on, still shows no confirmed public evidence of real-world adoption. It illustrates a general risk for other jurisdictions building similar voluntary certification schemes — legal architecture alone, without a market mandate, incentive, or transparency requirement on uptake, may not translate into actual use.