AI-enabled software delivery governance: an exploratory survey of practitioner readiness and workflow assurance
AI-enabled software delivery is expanding across coding, testing, review and operational support, but practitioner readiness and workflow assurance remain uneven. This exploratory cross-sectional survey analysed self-reports from 220 software practitioners recruited through non-probability convenience and selective peer-referral sampling. AI tools were used at least weekly by 70.9% of respondents, whereas 61.8% reported no formal training. On 0–100 deficit scales, awareness, common enablement, perceived routine implementation of guardrails and reported risk hotspots averaged 47.21, 46.48, 46.21 and 39.32, respectively; the secondary descriptive AI Governance Deficit Index averaged 44.80 (SD = 17.83). After adjustment for role family and professional experience, formal training was associated with a 6.74-point lower deficit score in the outcome excluding common enablement (95% CI [− 12.04, − 1.44]). Privacy/data-governance and robustness/safety familiarity were associated with 4.84- and 4.52-point lower deficit scores, respectively, in the outcome excluding awareness. Among 176 respondents who completed the optional written item, exploratory qualitative content analysis identified recurring workflow and evidence-capture gaps, data/tool/escalation ambiguity, and delivery-pressure or resource constraints. Because the study is cross-sectional, self-reported and based on non-probability sampling, the findings do not establish causal effects, verified control effectiveness, organisational maturity or improvements in software quality.
Authors
- Arun Karthik Dharmalingam
- T R Kalai Lakshmi
Institutions
- Sathyabama Institute of Science and Technology (IN)
Publication Details
- Journal
- Discover Computing
- Published
- 2026-11-20
- DOI
- https://doi.org/10.1007/s10791-026-10852-0
- Primary Topic
- Software Engineering Techniques and Practices
- Type
- article
- Field-Weighted Citation Impact
- 0.00