Errata for the 2025 DORA Report
This page lists errors and corrections to the 2025 State of AI-assisted Software Development report. To track revisions, report PDFs are stamped with a version number. The initial version of the 2025 report was v. 2025.1. The latest version is v. 2025.2.
Errata in v. 2025.2.p
The printed version of the 2025 report has the following errata:
p. 3 The page numbers in the table of contents are incorrect. They are all off by one, a classic error in computer science. The correct page numbers are:
- 04: Executive summary
- 09: Foreword
- 12: Understanding your software delivery performance
- 24: AI adoption and use
- 34: Exploring AI’s relationship to key outcomes
- 50: DORA AI Capabilities Model
- 66: Platform engineering
- 74: Value stream management
- 80: The AI mirror: How AI reflects and amplifies your organization’s true capabilities
- 90: Metrics frameworks
- 96: Final thoughts: From insight to action
- 98: Acknowledgements
- 100: Authors
- 104: Demographics and firmographics
- 114: Methodology
- 132: Our research model and its theory
- 138: Next steps
- 139: Appendix
p. 71 In the fourth paragraph, “stabilty” should be “instability.”
- “The increase in performance in spite of the increase in stability, suggests a form of risk compensation where the platform makes it fast and cheap to recover…” should be “The increase in performance in spite of the increase in instability, suggests a form of risk compensation where the platform makes it fast and cheap to recover…”
Errata in v. 2025.2
p. 51 Figure 32 has an incorrect caption. The correct caption is “Figure 32: A clear and communicated AI stance determines AI’s impact on organizational performance”
p. 70 In the fourth paragraph, “stabilty” should be “instability.”
- “The increase in performance in spite of the increase in stability, suggests a form of risk compensation where the platform makes it fast and cheap to recover…” should be “The increase in performance in spite of the increase in instability, suggests a form of risk compensation where the platform makes it fast and cheap to recover…”
Errata in v. 2025.1
p. 9 There is an extra “deployments” in “This shift helped enable 136,000 deployments deployments per day in 2015” [Corrected in v. 2025.2]
p. 20 Figure 12: Distribution of deployment frequency responses from the 2025 DORA Survey. The deployment frequency in the second row should be “Between once per month and once every six months”. [Corrected in v. 2025.2]
p. 44 Figure 29: Estimated effect of AI adoption score scaled on key outcomes. “Existential oonnection” on the Y-axis should be “Existential connection” [Corrected in v. 2025.2]
p. 46 Figure 30: Ownership. The figure does not have a key explaining the meaning of the colors. [Corrected in v. 2025.2]
- Light blue
is No AI. - Dark purple
is AI.
p. 51 Figure 32 has an incorrect caption. The correct caption is “Figure 32: A clear and communicated AI stance determines AI’s impact on organizational performance”
p. 52 Figure 34 has an incorrect title and caption. [Corrected in v. 2025.2]
- The correct title is “A clear and communicated AI stance moderates AI’s impact on software delivery throughput”
- The correct caption is “Figure 34: A clear and communicated AI stance moderates AI’s impact on software delivery throughput”
p. 57 Figure 38 has an incorrect title and caption. [Corrected in v. 2025.2]
- The correct title is “Version control commit frequency moderates AI’s impact on individual effectiveness”
- The correct caption is “Figure 38: Version control commit frequency moderates AI’s impact on individual effectiveness”
p. 70 In the fourth paragraph, “stabilty” should be “instability.”
- “The increase in performance in spite of the increase in stability, suggests a form of risk compensation where the platform makes it fast and cheap to recover…” should be “The increase in performance in spite of the increase in instability, suggests a form of risk compensation where the platform makes it fast and cheap to recover…”
p. 97 Footnote 1 should link to https://dora.dev/research/2025/questions.
[Corrected in v. 2025.2]
p. 100 “He has studied the deployment of intelligent technologies since 2011.” has an unnecessary space at the beginning of the sentence. [Corrected in v. 2025.2]
p. 107 Figure 61: Programming language usage. There are a few errors in the list of programming languages: [Corrected in v. 2025.2]
- The list includes two entries for “C”. The first should be “C#” (18.5%) and the second should be “C” (3.9%).
- It lists “C+” as a language. It should be “C++” (14.5%).
p. 109 Figure 65 has an incorrect title and caption, “unavailability” is misspelled.
[Corrected in v. 2025.2]
- The correct title is “Impact of primary service unavailability”
- The correct caption is “Figure 65: Impact of primary service unavailability”
p. 115 The description for one of the core analysis variables, ai_adoption, is incorrect. It should state: “ai_adoption = a factor made of three indicators”. [Corrected in v. 2025.2]