AI is no longer experimental; it’s operational. It determines who qualifies for mortgages, how insurance is priced, what health plans are covered, and which resumes are approved by automated filters. For CLOs, CDOs, and enterprise leaders, the urgency isn’t simply “how do we use AI?” but rather “how do we ensure AI works responsibly, without eroding trust, equity, or competition?”
That tension was front and center in The Hidden Impact of Responsible AI, a conversation hosted by Merav Yuravlivker, Chief Learning Officer of Data Society Group and Co-Founder of Data Society, and featuring Carey Nadeau, Co-CEO of LOOP Insurance.
Carey brings the lens of a builder and operator. After a career modeling cities, transportation, and equity at institutions like MIT and Brookings, she co-founded LOOP to reimagine how car insurance could be priced more fairly. Her work includes encoding the Living Wage, developing the first commercialized geospatial machine learning models to predict car crashes, and using public data to drive accountability in industries where opacity is the norm.
Together, their conversation explored a core question: how do we apply AI with clarity and purpose to reshape public safety, equity, and policy, beyond the hype and headlines?
Below, we share six takeaways from Carey’s insights, grounded in Merav’s framing and questions.
MUST READ: From Data to Decisions: How West Point Builds AI, Data, and Decision Literacy at Every Level
1. Anchor AI in Real-World Impact
AI isn’t valuable just because it’s advanced; it’s valuable when it tangibly improves lives. Carey recalls working with the City of Chicago on one of the first geospatial machine learning models to predict where car crashes were most likely to occur. That model didn’t live in a lab. It directly shaped how the city invested in new stoplights, resurfaced roads, and extended crosswalk signals near schools and elder communities.
The lesson: data becomes powerful when it translates into outcomes people can feel in their daily lives. It’s the difference between talking about “predictive modeling” and ensuring a grandmother makes it safely across the street.
“I’m not asking questions just to ask questions. I’m building for a specific outcome I care about, something that changes lives in my community.”– Carey Nadeau
Takeaway: AI without impact is wasted effort.
Actionable Insight: Frame every initiative with a “so what?” If you can’t clearly articulate who benefits and how, pause before investing further.
2. Keep Data Open Enough to Compete
The early open data movement demonstrated what was possible when government datasets were made public: new services, new companies, and improved accountability. But today, “open” has been hollowed out. “OpenAI” is anything but open. A handful of private companies hold and guard the majority of the data that fuels modern AI.
For Carey, that’s dangerous. Without openness, innovation shrinks, monopolies form, and bias calcifies. Her provocative solution: rebrand openness as “naked AI.” Transparency, competition, and diversity of perspective, she argues, aren’t just moral imperatives; they’re long-term business advantages.
“Building in the open increases competition. It invites more people, more perspectives, and creates better products.”– Carey Nadeau
MUST READ: Scaling AI with Governance: Practical Advice from Lockheed Martin’s Mike Baylor
3. Don’t Rely on Watchdogs Alone
It’s tempting to call for independent auditing boards or government committees to monitor AI systems. But Carey warns that these mechanisms are often slow, underfunded, or captured by the very industries they’re meant to police. Certification stamps look good, but too often they’re shallow.
Instead, she argues, the real check on concentrated power is competition. History shows that monopolies, whether railroads, oil, or telecommunications, were only broken when competition was restored. AI, she suggests, will be no different.
“What checks unchecked power? More competition. It’s how the American economy has always recovered from monopolies.”– Carey Nadeau
Takeaway: Oversight is necessary, but competition is the real safeguard.
Actionable Insight: When evaluating AI tools, ask: Does this vendor expand my choices or narrow them? Choose partners who thrive on competition, not those who seek to lock you in.
4. Reimagine Data Privacy for the Real World
For many leaders, “data privacy” is still framed as a goal. For Carey, it’s already a myth. Every click, location ping, and purchase is tracked. Every feed is optimized to nudge behavior. Ads chase us across the internet. The cat is out of the bag.
Rather than fighting to restore what can’t be reclaimed, she suggests reframing the battle: demand value for what is taken. If privacy is gone, then the exchange must deliver safety, equity, or opportunity in return.
“I’d give up data privacy for more justice, more equity, more health. The problem isn’t that we give up privacy, it’s that we’re not getting enough back for the exploitation.”– Carey Nadeau
Takeaway: Privacy battles are less useful than demanding value in return.
Actionable Insight: Shift the conversation with employees and customers: what do they gain in exchange for their data? Be explicit about benefits, not just about permissions.
5. Navigate the AI Bubble with Caution
Venture capital is pouring billions into a small circle of AI companies, accelerating growth but narrowing diversity. Founders from underrepresented backgrounds, already receiving less than 1% of venture funding, are now virtually shut out. The result? AI systems increasingly reflect the perspectives and incentives of a single demographic.
Carey likens this to creating an AI monoculture: a system that could one day rival human intelligence but reflect only one worldview. The risk isn’t just ethical, it’s strategic. Homogeneous perspectives lead to blind spots, brittle systems, and missed opportunities.
“What if AI only has one point of view? I don’t want to live in that world. Diversity of ideas is where real innovation comes from.”–Carey Nadeau
Takeaway: The AI gold rush risks narrowing, not expanding, human creativity.
Actionable Insight: Ask your vendors about who builds and trains their systems. Look for diversity in datasets, development teams, and leadership, not as a checkbox, but as a safeguard for resilience.
6. Remember: Human Intent Still Matters
Despite the hype, AI is still a tool. It reflects the intent of the people who design, deploy, and use it. For Carey, that’s both empowering and sobering. Leaders still choose whether AI amplifies creativity or automates exploitation, whether it frees workers for meaningful work or simply cuts costs.
“It’s not just what tools we use. It’s how we use them, why we use them, and the intent we bring to them.” –Carey Nadeau
Takeaway: Technology only amplifies the intent behind it.
Actionable Insight: Pair AI adoption with leadership training that emphasizes ethics, purpose, and transparency. Equip teams to ask not just can we? but should we?
The Next Chess Move
AI’s trajectory isn’t inevitable. As Carey reminded us, leaders can choose openness over secrecy, competition over monopoly, equity over exploitation. Each decision is a chess move that shapes the future. The board is set, but the next move is ours.
Watch the replay of The Hidden Impact of Responsible AI here
FAQ: Responsible AI, Data, and Transparency
Carey suggested “naked AI” as a rebrand for what used to be called “open AI” or “open data.” The idea is to build transparently, in public view, where anyone can challenge, compete, and improve the work. Naked AI invites accountability and avoids the risks of monopolized, closed systems.
