I write from my own work, research, and experience. AI may help with ideation, research, outlining, drafting, or editing. I take responsibility for every claim I publish, the evidence supporting it, and my interpretation of that evidence.
Before publication, I review the article myself and check its factual claims. I support technical and research claims with outside sources I have read myself. For a cited paper, that includes examining the methods, results, and limitations relevant to the claim. I use AI summaries to find material to investigate; I still do that investigation and decide what the source supports.
I distinguish what a source establishes from conclusions I draw from it. When the evidence leaves a question unresolved, I say so. Follow-up research is part of checking an interpretation, especially when another source or a reader gives me reason to doubt it.
Accounts of my work must describe things I actually did or observed. Where possible, I provide code, records, or reproducible results that readers can inspect. I label hypothetical examples and proposed designs, and distinguish them from completed work and measured results. An illustration can explain my reasoning; it cannot serve as evidence of experience I haven’t had.
Accuracy determines what I am willing to publish. I check claims as an article develops and review them again before publication. When AI has substantially helped draft an article, I disclose that assistance on the article.
If I discover an error, I correct it and add a dated note when the correction changes a claim or conclusion. Publication dates record when articles first go live. Substantive revisions receive an update date.
If you find an unsupported claim, a misleading interpretation, or an error, contact me on LinkedIn with the passage and any relevant evidence. I will investigate and update the article when a correction is warranted.
AI assisted with drafting and editing this policy.