Christine "Ink" Whitmarsh
12 November 2026
40m 7s
Your Data Is Going Somewhere: A HumanPrint Interview with Apostolos Stamenos
00:00
40:07

Christine "Ink" Whitmarsh
12 November 2026
40m 7s
00:00
40:07
That conversation you’re having with AI? It’s also data going somewhere.
In this episode of HumanPrint, I talk with Apostolos Stamenos, an award-winning former federal technologist and founder of ouroborOS Consulting, about what responsible AI use looks like when privacy, accuracy, and data governance actually have consequences.
During the COVID-19 response, Apostolos worked with public-facing hospital data at HHS, where bad numbers could eventually reach Congress, the media, and the public. That experience taught him to think about AI systems the same way he thinks about any other data infrastructure: What can fail? What can fail silently? Where is the information going? And who will notice if something goes wrong?
We talk about why a conversational interface can make people forget they’re still submitting data to a company’s servers, the difference between local and proprietary AI models, what organizations may be giving away when employees paste information into ChatGPT or Claude, and why privacy is about much more than removing somebody’s name from a document.
We also get into AI-generated fraud and manipulation, the difference between what a chatbot “remembers” and what the company behind it may retain, and why supposedly harmless platform features can also double as data annotation systems.
Apostolos’ starting question for organizations considering AI is refreshingly simple: What are the human stakes? The reputational stakes? The financial stakes?
Because sometimes the responsible answer really is: don’t paste that into the chatbot.
Learn more and connect with Apostolos: https://substack.com/@astamenos