What DigComp 3.0 Is and Why Lithuania Named It the National AI Literacy Standard
DigComp 3.0 appears in Lithuania's national AI strategy, in EU AI Act guidance, and in Article 4 enforcement discussions. Most public sector professionals who encounter the term do not know what the framework actually contains.
This post explains what DigComp 3.0 is, what changed from the previous version, and why Lithuania named it the national AI literacy standard.
Key points
- DigComp 3.0 is the European Commission's current digital competence framework (JRC144121, November 2025). It has 5 competence areas, 21 competences, 8 proficiency levels, and over 500 learning outcomes.
- DigComp 3.0 is the first EU framework to integrate AI competence across all 21 competences. Eighty-two percent of its competence statements carry AI relevance.
- Lithuania's National AI Strategy 2026-2035 names DigComp 3.0 as the recommended framework for national AI literacy programmes.
- The Digital Omnibus (July 2026) states that organisations can use European competence frameworks including DigComp to satisfy Article 4. Using DigComp 3.0 for compliance documentation is directly supported by the legislative record.
- A DigComp 3.0 competence report shows what a learner can do at a specific proficiency level. A generic course certificate does not.
What DigComp 3.0 is
DigComp 3.0 is the European Commission's current framework for describing digital competence. Published by the Joint Research Centre (JRC) in November 2025 (JRC144121), it was developed with the Directorate-General for Employment, Social Affairs and Inclusion.
The framework covers 5 competence areas, 21 individual competences, and 8 proficiency levels. It includes over 500 learning outcomes: observable descriptions of what competence looks like in practice at each level.
DigComp 3.0 is a competence descriptor, not a qualification framework. Completing DigComp 3.0-aligned training does not produce a formal degree or credit equivalent. It produces a competence profile: a structured record of which competences a person has demonstrated, at which proficiency level, for what context.
That distinction matters for Article 4. What counts as compliance evidence is documentation that maps training outcomes to the specific competences and contexts each role requires. A competence profile does that. A generic course certificate does not. For a full breakdown of what Article 4 compliance documentation must contain, see What EU AI Act Article 4 Requires of Lithuanian Public Sector Bodies.
A brief history of DigComp
DigComp 1.0 was published in 2013. It was the EU's first attempt to describe digital competence in a way that member states could use for policy and curriculum design. The original framework covered five areas with 21 competences.
Version 2.2 followed in 2022. It added examples covering AI tools and data handling, but AI remained illustrative rather than central. The competence statements themselves did not change to reflect how fully AI tools had entered professional practice.
With DigComp 3.0, AI competence is built into the framework's competence statements for the first time, across all five areas. The update recognises that digital competence and AI competence cannot be separated in the professional context of 2026.
The five competence areas
DigComp 3.0 organises competence around five areas. Each area contains a set of individual competences, and each competence has eight proficiency levels.
1. Information and data literacy. Finding, evaluating, and managing digital information. This area now explicitly covers how AI systems generate, rank, and present information, and how to critically assess AI-generated content.
2. Communication and collaboration in digital environments. Engaging with others through digital channels, sharing knowledge, and collaborating in digital spaces. For public sector AI deployers, this area covers how AI tools mediate institutional communication and citizen engagement.
3. Digital content creation. Producing, editing, and managing digital content across formats. In DigComp 3.0, this area carries the highest concentration of AI-explicit competences. The JRC framework reflects the reality that AI content generation tools now shape many professional workflows.
4. Safety. Protecting devices, data, privacy, and wellbeing in digital contexts. This covers cybersecurity, data protection, and the specific risks AI systems introduce for individuals and organisations.
5. Problem solving. Using digital tools to address problems and identify needs. For public sector deployers, this includes knowing when to apply an AI system and how to respond when AI output is incorrect.
The five areas are equally applicable to a Lithuanian civil servant as to any EU professional using AI systems at work. DigComp 3.0 is jurisdiction-neutral by design. Its application to the Lithuanian public sector is defined by Lithuania's National AI Strategy 2026-2035.
The eight proficiency levels
DigComp 3.0 uses eight proficiency levels, grouped into four bands.
Foundation (Levels 1 to 2). The learner can perform basic digital tasks with guidance. At this level, AI awareness is limited to recognising that AI systems exist and are in use.
Intermediate (Levels 3 to 4). The learner works independently with digital tools in familiar and some unfamiliar contexts. This is the level at which most Lithuanian public sector professionals should be operating. A front-line officer who uses an AI case management tool and understands its limitations is working at Intermediate level.
Advanced (Levels 5 to 6). The learner adapts their digital practice to complex and new situations. They can guide others and critically evaluate digital tools. A digital transformation lead or AI programme manager would typically need to demonstrate Advanced competence.
Highly Specialised (Levels 7 to 8). The learner contributes to developing digital practice in their field. For most public sector roles, these levels are above what Article 4 compliance training targets.
The proficiency level distinction is what separates a defensible Article 4 record from a generic one. The obligation requires measures appropriate to each person's role and context. A record that does not name a proficiency level cannot demonstrate that the training was proportionate to the role.
What changed from DigComp 2.2 to 3.0
The structure of DigComp 3.0 matches version 2.2: five areas, the same 21 competences, the same eight levels. What changed is the content of those competences, and the change is substantial.
A quantitative analysis of DigComp 3.0's competence statements found that 82 percent carry some AI relevance. Fourteen percent are AI-explicit, describing competences that are directly about AI systems. The remaining 68 percent are AI-implicit, covering skills where AI tools have become a central part of how work is done.
This is the first EU competence framework to integrate AI systematically across all competences, rather than as a separate module.
Area 3 (Digital Content Creation) has the highest concentration of AI-explicit statements, at 45 percent of that area's total. This reflects how central AI content generation tools have become to professional digital work.
Other significant updates include expanded cybersecurity competences, digital rights and responsibilities, and competences addressing misinformation and synthetic content. All of these are directly relevant to public sector AI deployers.
In version 2.2, AI appeared mainly in illustrative examples added to existing competence statements. In 3.0, AI is part of what the competences describe. That difference is why DigComp 3.0, not 2.2, is what Lithuania's national AI strategy references.
Why Lithuania named DigComp 3.0 the national AI literacy standard
Lithuania's National AI Strategy 2026-2035 names DigComp 3.0 as the recommended framework for national AI literacy programmes. This is a specific policy decision by the Lithuanian Ministry of Economy and Innovation.
The GovAI Competence Center, established in early 2026, implements that decision. It provides DigComp 3.0 competence reports to government institutions and has been training Lithuanian public sector staff since launch. By 2025, approximately 6,000 Lithuanian public sector employees had received AI skills training, with DigComp 3.0 as the measurement standard.
Two things make DigComp 3.0 the correct framework for an Article 4 compliance context, not just the nationally recommended one.
First, the Digital Omnibus (Regulation (EU) 2026/1744, July 2026) is explicit. For Article 4 compliance, the Commission and member states may take into account European competence frameworks including DigComp. That reference is in the legislative record. Using DigComp 3.0 to document Article 4 compliance is not a creative interpretation of the law.
Second, DigComp 3.0 is compatible with the OECD AI Literacy framework: all OECD AI literacy competences are present within DigComp 3.0. This cross-framework compatibility means DigComp 3.0 documentation satisfies both EU and international reference standards.
What this means for an institution
For a Lithuanian public sector institution, DigComp 3.0 is now the national AI literacy standard. The practical question is not whether to use it, but how.
Three steps follow from this.
Map each role to the DigComp competences relevant to the AI systems that role uses. A procurement officer evaluating AI vendors needs competences from Area 1 (information and data literacy) and Area 5 (problem solving). A front-line case officer using an AI support tool needs competences from Area 4 (safety) and Area 2 (communication and collaboration). The mapping must be specific to the AI systems your institution has deployed.
Identify the proficiency level appropriate to each role. A civil servant who uses AI tools only in standardised workflows may need Intermediate Level 3. One who guides institution-wide AI adoption needs Advanced Level 5 or 6. The same content at the same level does not serve both roles.
Produce a competence record that names competences and levels, not just completion dates. An attendance log names no competence. A DigComp 3.0 competence report names the competences developed, the proficiency level reached, and the assessment basis. That is what Article 4 compliance documentation looks like in practice.
How a DigComp 3.0-aligned programme differs from a generic AI course
Many training programmes now claim DigComp 3.0 alignment. The claim is meaningful only if the programme can demonstrate which specific competences it develops, at what proficiency level, assessed through what method.
A genuinely DigComp 3.0-aligned programme does three things:
- Tags every piece of content to one or more specific DigComp 3.0 competences
- Assesses learners against those specific competences at the proficiency level the role requires
- Produces a competence report naming the competences demonstrated and the proficiency level reached
A course that covers AI concepts, includes a quiz, and awards a certificate may be useful. It is not DigComp 3.0-aligned unless it can answer: which competences did this learner develop, to what level, as confirmed by this assessment?
The difference is evidential. Article 4 requires measures appropriate to each person's role and AI use context. A generic completion record cannot demonstrate appropriateness. A competence report mapped to specific DigComp areas and levels can.
Kelias maps every module, assessment, and case study to specific DigComp 3.0 competences and proficiency levels. The full competence mapping is available at kelias.tech/curriculum.
Kelias is pre-commercial and not yet open for enrolment. To join the waitlist or register for the free 30-day institutional pilot, visit kelias.tech/organisations.
Ikpong Joseph Alexander holds an MSc in Artificial Intelligence and is the founder of Kelias. The primary source for DigComp 3.0 is the JRC European Digital Competence Framework (JRC144121), available from the JRC DigComp 3.0 page. This post cites the Digital Omnibus as Regulation (EU) 2026/1744, in force 27 July 2026.