Overview

AI4SE1DK (Human-Centred Adoption of Artificial Intelligence for Software Engineering) in Denmark is a national four-year research programme positioning Denmark as a global reference for trustworthy, human-centred, and sustainable adoption of AI in software engineering. The programme closes the current adoption gap while strengthening the competitiveness of Danish software organisations.

Aim

One of the aims of AI4SE1DK is to identify and investigate solutions for assessment of AI-generated software in terms of

• security;
• sustainability;
• and productivity.

Structure

WP4 (AI Software Assessment) is structured into three tasks:

• T4.1 Software assessment methods: Identify state-of-the art solutions for assessment of AI generated code in terms of security (quality/trustworthiness), sustainability (energy efficiency), and productivity.
• T4.2 Scalable assessment methods: Identifying scalable and sustainable code assessment solutions, where human efforts and resource consumptions are reduced and can scale with the speed of code production, without impacting the quality of the assessment.
• T4.3: Automated software assessment: Identify methods that combine classical automated code analysis such as static analysis, dynamic analysis, formal verification, and testing with novel AI-based solutions to increase quality, sustainability, productivity, and trustworthiness.

Deliverables

The learnings from the three WP4 tasks are summarized in this page, structured as these deliverables:

• D4.1: Snapshot of this page as of September 2026. The focus in this period is on state of the art solutions (outcome of T4.1).
• D4.2: Snapshot of this page as of April 2027. The focus in this second period is on novel approaches developed within the project (outcome of T4.2).
• D4.3: Snapshot of this page as of September 2028. The focus in this third and last period is in combining the novel contributions of the project and state of the art solutions (outcome of T4.3).

Rest of This Page

The rest of the page, contains our main findings, structured around key questions relevant to AI software development.