Puru Trivedi, Senior Vice President at Meridian International Center, argues that Washington’s influence industry may be the last business AI cannot disrupt — because it runs on human trust, relationships, and judgment that no model can replicate. Host Vivek Kumar connects that thesis to what Teamforce AI sees on the factory floor every day: as AI systems become more embedded in industrial operations, the human signals that prevent catastrophic failures become more valuable, not less.
The Town That Runs on Trust
Every few months, a senior AWS executive meets with state governors, mayors, and utility officials across the United States. The conversations are not about cloud architecture. They are about water usage, electricity consumption, land use, and zoning. The topic is how to change the perception of what data centers mean in local communities.
Puru Trivedi has a front-row seat to those conversations. As the SVP leading all non-government work at Meridian International Center, a 65-year-old public diplomacy institution ten blocks north of the White House, Trivedi works with roughly 120 member companies including Boeing, Chevron, Delta, Visa, and the major technology platforms. His perspective on AI is shaped by what he hears across that network every week.
His core observation is blunt: Washington may be the one industry AI cannot disrupt.
The reason is structural. Influence in Washington operates on what Trivedi calls “hand-to-hand combat.” Who you know. Who returns your call. Who trusts you enough to take a meeting they do not have time for. These are relationship outputs that cannot be generated, accelerated, or faked by a language model or an automation layer.
The Attention Economy Has a Factory Floor Equivalent
Trivedi’s point about Washington applies well beyond the Beltway. There is a competition for attention like there has never been before, and AI is accelerating it. In Washington, that competition plays out in lobbying, legislation, and media cycles. On the factory floor, it plays out in a different but equally consequential way: which signals get heard, which hazards get reported, and which corrective actions actually reach the people they need to reach.
Most manufacturers have more data than they can act on. Sensors stream. Dashboards update. Compliance platforms log. But the signal that prevents the next serious injury or quality escape is often not in the data system at all. It is in the head of a frontline worker who saw something, knows something, or suspects something — and has no trusted, frictionless way to surface it.
This is the human layer. AI cannot generate it. AI cannot observe it from a camera feed. And as more AI-driven automation enters the plant, the gap between what the system sees and what the human knows is growing, not shrinking.
Educating the People Who Write the Rules
Meridian runs a program called Tech Labs that brings congressional and Senate staff to learn about emerging technology in a nonpartisan, non-corporate setting. The goal is to ensure the people drafting AI legislation understand the technology they are regulating. A separate track runs for the executive branch and the administration.
Trivedi notes that there are knowledgeable, dedicated public servants in Washington working on AI policy. But the AI conversation is getting crowded out by more immediate concerns: the economy, geopolitics, midterm elections, and kitchen-table issues that have dominated since the 2024 election cycle. The result is that the regulatory landscape for AI in manufacturing and industry is being shaped by people who care but are stretched thin.
For self-insured manufacturers deploying AI on the factory floor, this matters. The regulatory environment is forming now. Companies that can demonstrate accountability — not just compliance documents, but proof that interventions reached the workforce, that corrective actions were verified, that the human-AI handoff is working — will be better positioned as the rules take shape.
When Humans and AI Share the Floor, Risk Does Not Go Down Automatically
Teamforce AI Co-Founder and CEO Vivek Kumar frames the emerging risk landscape this way: when humans interact with AI systems that are not fully deterministic — systems that have, as he puts it, “a little bit of their own mind” — you do not automatically get safer operations. You get new categories of risk and uncertainty that did not exist when the process was fully manual or fully mechanical.
The assumption in many boardrooms is that more AI equals less human error. The reality on the floor is more nuanced. AI introduces new failure modes: operators deferring to a system that is wrong, alert fatigue from too many automated flags, workarounds that develop when the AI-driven process does not match the physical reality of the line. These are human signals. They do not appear in the sensor data. They appear in what the workforce says — if anyone is listening.
Cross-Cultural Fluency as Operational Advantage
Trivedi is also the author of My Fourth Culture, a memoir that traces his path from Boston to Kolkata to Saudi Arabia to Washington. His family’s story spans the 1947 partition, three continents, and careers in finance, diplomacy, and government. He wrote the book during the pandemic as a way to preserve memories of his mother, who passed in 2018, and did not realize until later that it would become a record for his own child.
His thesis on what comes next is relevant for any manufacturer operating across geographies and cultures: cross-cultural fluency — the ability to move between contexts, read rooms, and build trust across differences — is going to be a significant competitive advantage precisely because AI commoditizes everything else. When the output looks the same, the relationship is the differentiator.
On the factory floor, this translates directly. Plants with multilingual workforces, mixed shifts, and diverse cultural norms need more than a translated poster. They need systems that meet workers where they are, in their language, in their context, without requiring them to navigate a corporate reporting structure that was not designed for them.
Why This Matters for the COO and CFO
The policy landscape around AI in manufacturing is forming. Meridian is helping to shape it at the congressional level. The companies that will navigate that landscape most effectively are not the ones with the most sophisticated AI deployments. They are the ones that can prove the human layer is working: that frontline signals are being captured, that fixes are verified, and that the workforce is an active participant in safety and operational outcomes rather than a passive recipient of automated instructions.
AI is a tool. Trust is infrastructure. And infrastructure is what survives regulation.
Watch the Full Episode
Hear the full conversation with Puru Trivedi on AI, Washington, cultural diplomacy, and why human trust is the layer that cannot be automated.
All proceeds from Puru’s book My Fourth Culture go to the Dana-Farber Cancer Institute and St. Jude’s Hospital.
Connect with Puru Trivedi on LinkedIn.