In the AI: Trust but Verify podcast, our host, Alec Crawford (@alec06830), Founder and CEO of Verapath (www.verapath.com), interviews guests about how they are using AI in business, where you can trust AI, and where you need to put up guard rails. Podcast production and sound engineering by Troutman Street Audio. You can find them on LinkedIn.

AI: Trust but Verify — with Dominick Romano

About the Guest

Dominick “Dom” Romano is the founder and CEO of Drainpipe.io, an AI company focused on making AI systems trustworthy enough for regulated, high-stakes environments. Dom’s background spans video game development, casino gaming, advertising, real-time routing for hazardous payloads, mainframe engineering for IBM z/OS and banking systems, real-time transactions, and observability. Today, Drainpipe.io works with major manufacturers in Germany across pharmaceuticals, automotive, and chemical manufacturing, helping them deploy AI systems where the inputs and outputs must be verifiable—especially when AI touches critical regulatory data such as pharmaceutical dossiers submitted to health authorities.

Top 5 Takeaways

AI adoption in regulated industries requires trust, not just capability. Dom emphasizes that when AI touches critical workflows—especially in pharmaceuticals, compliance, or regulatory submissions—organizations need confidence that both the data going in and the AI-generated output are legitimate and trustworthy.
AI can scale mistakes, bias, and discrimination. The conversation highlights how AI systems used in decisions such as lending or hiring can create large-scale harm if bias or unlawful discrimination goes undetected—particularly when companies cannot prove the model is not using impermissible factors.
Cybersecurity has to come before “cool” AI features. Dom and Alec discuss how rapidly adopted AI tools can create serious security risks when they go viral without proper cybersecurity foundations. The OpenClaw example is used as a warning about software that provides value but fails to account for security from the start.
AI is becoming a geopolitical and cultural force. Dom raises concerns about “digital colonialism,” where countries in the Global South may become increasingly shaped by Western AI models that do not represent their own cultures, languages, or values.
The near-term AI risks may be more urgent than distant sci-fi scenarios. While existential risk is discussed, Dom argues that immediate threats—deepfakes, multimodal AI on devices, AI in armed conflict, and rapidly expanding cybersecurity vulnerabilities—deserve serious attention right now.

People and Organizations Mentioned

Dominick “Dom” Romano — Founder & CEO, Drainpipe.io
Website: https://drainpipe.io
LinkedIn: linkedin.com/in/domromano
X: https://x.com/dromanocpm
Dom specifically points listeners to LinkedIn, X under the handle @drobmanocpm, and
Alec Crawford — Founder & CEO, Verapath; host of the conversation linkedin.com/in/aleccrawford
Verapath — Secure AI platform for financial institutions https://verapath.com
Full Sail University — Dom’s educational background in video game development
https://www.fullsail.edu
IBM z/OS — Mainframe platform referenced in Dom’s background
https://www.ibm.com/products/zos
EU AI Act — European AI regulation discussed in the episode https://artificialintelligenceact.eu
JP Morgan / JPMorgan Chase — Mentioned in the discussion of AI, attrition, and jobs
https://www.jpmorganchase.com
Anthropic — Mentioned in the discussion of AI existential risk https://www.anthropic.com
Ray Kurzweil — Mentioned as one of the people concerned about AI risk https://www.kurzweilai.net
Elon Musk — Mentioned in the discussion of AI risk https://x.com/elonmusk
Marcus Hutter — AI superintelligence researcher mentioned by Alec https://www.hutter1.net
Doraemon — Referenced as an example of Japan’s more positive cultural association with AI
https://dora-world.com
OpenClaw — Referenced as an example of a viral software/AI-adjacent tool with serious cybersecurity concerns
Mythos — Referenced in the discussion of emerging cybersecurity threats and online infrastructure risk