Project Details
Description
This German-Australian research project examines how organizations under high regulatory scrutiny manage tensions between normative responsibility in AI systems and economic performance. As global efforts toward Responsible AI (RAI) intensify, organizations face growing challenges in adapting AI practices (Fraser et al., 2024; Högberg, 2024). The EU’s AI Act mandates strict transparency, fairness, and accountability, while Australia takes a more principles-based “guardrails” approach (Fraser et al., 2024). In 2024–25, the Australian Federal Government released a discussion paper on responsible AI, inviting feedback from industry and academia.
These developments respond to concerns about opaque, biased algorithms and
broader societal risks (Kuss & Meske, 2025; Meske et al., 2022; Abedin, 2022). At the same time, AI promises innovation, efficiency, and business transformation (Mikalef & Gupta, 2021). However, heavy compliance requirements can slow agile
experimentation and increase costs, exacerbating tensions between governance and performance (Berente et al., 2021; Högberg, 2024). The project draws on
organizational ambidexterity, the ability to balance exploitation (efficiency) with
exploration (innovation) (Andriopoulos & Lewis, 2009). In the AI context, this involves aligning compliance with innovation goals. Although dynamic capabilities enable adaptation to AI disruptions (Mikalef & Gupta, 2021; Hillebrand et al., 2025), the interaction between governance and these capabilities remains underexplored (Berente et al., 2021).
The project consists of five work packages: WP1 (Months 1–3): Systematic literature review on RAI, governance, ambidexterity, and sociotechnical design. A shared conceptual model will define key terms like “responsibility,” “compliance,” and “innovation.” WP2 (Months 4–5): Document analysis of EU and Australian regulatory texts, corporate guidelines, and existing research (including 250 Australian government submissions). Interview guides will be designed and tested for crossnational use. WP3 (Months 6–11): 30–40 interviews with AI practitioners, compliance officers, and innovation leads in Germany and Australia, using networks like HUMAINE and Billigence. WP4 (Months 12–15): Comparative, theory-driven analysis to develop a cross-national RAI framework aligned with innovation and regulatory goals. WP5 (Months 12–24): Dissemination through two academic articles, two practice-oriented outputs (including a policy brief), conference presentations, and planning for future collaboration.
The project aims to advance understanding of how firms implement RAI while
sustaining innovation. Key outputs include a conceptual model, validated methods, joint publications, Master theses, and a long-term research partnership.
These developments respond to concerns about opaque, biased algorithms and
broader societal risks (Kuss & Meske, 2025; Meske et al., 2022; Abedin, 2022). At the same time, AI promises innovation, efficiency, and business transformation (Mikalef & Gupta, 2021). However, heavy compliance requirements can slow agile
experimentation and increase costs, exacerbating tensions between governance and performance (Berente et al., 2021; Högberg, 2024). The project draws on
organizational ambidexterity, the ability to balance exploitation (efficiency) with
exploration (innovation) (Andriopoulos & Lewis, 2009). In the AI context, this involves aligning compliance with innovation goals. Although dynamic capabilities enable adaptation to AI disruptions (Mikalef & Gupta, 2021; Hillebrand et al., 2025), the interaction between governance and these capabilities remains underexplored (Berente et al., 2021).
The project consists of five work packages: WP1 (Months 1–3): Systematic literature review on RAI, governance, ambidexterity, and sociotechnical design. A shared conceptual model will define key terms like “responsibility,” “compliance,” and “innovation.” WP2 (Months 4–5): Document analysis of EU and Australian regulatory texts, corporate guidelines, and existing research (including 250 Australian government submissions). Interview guides will be designed and tested for crossnational use. WP3 (Months 6–11): 30–40 interviews with AI practitioners, compliance officers, and innovation leads in Germany and Australia, using networks like HUMAINE and Billigence. WP4 (Months 12–15): Comparative, theory-driven analysis to develop a cross-national RAI framework aligned with innovation and regulatory goals. WP5 (Months 12–24): Dissemination through two academic articles, two practice-oriented outputs (including a policy brief), conference presentations, and planning for future collaboration.
The project aims to advance understanding of how firms implement RAI while
sustaining innovation. Key outputs include a conceptual model, validated methods, joint publications, Master theses, and a long-term research partnership.
| Acronym | DAAD25 |
|---|---|
| Status | Active |
| Effective start/end date | 1/01/26 → 31/12/27 |