Guardians of AI: Jeff Radwell of Camouflet On How AI Leaders Are Keeping AI Safe, Ethical, Responsible, and True
An Interview With Yitzi Weiner
Explainability and Justification: If an AI system can’t explain its decisions, it shouldn’t be making them. Look at Australia’s RoboDebt disaster, where an automated system falsely calculated welfare overpayments, devastating thousands of people financially.
As AI technology rapidly advances, ensuring its responsible development and deployment has become more critical than ever. How are today’s AI leaders addressing safety, fairness, and accountability in AI systems? What practices are they implementing to maintain transparency and align AI with human values? To address these questions, we had the pleasure of interviewing Jeff Radwell.
Jeff Radwell, Ph.D., MBA, is an author, entrepreneur, and scientist whose expertise spans molecular genetics, computational biology, and AI-driven innovation. His achievements in the life and mathematical sciences have earned him recognition in academia and industry, including professorships at Imperial College London and NYU Grossman School of Medicine. As a scientific innovator, he holds multiple patents, and Hhs work continues to shape modern science, advancing how AI and biomedical research intersect to drive meaningful discoveries.
Thank you so much for your time! I know that you are a very busy person. Before we dive in, our readers would love to “get to know you” a bit better. Can you tell us a bit about your ‘backstory’ and how you got started?
My path to founding Camouflet wasn’t exactly traditional. I started in molecular biology, working in machine learning for public health research. During the COVID-19 pandemic, I was part of a team at NYU Grossman School of Medicine that built the data infrastructure to help deepen our understanding of the virus. That work led to the development of the first computable phenotype for Long COVID, a way to systematically define and identify the condition using real-world clinical data.
That experience reinforced something I had always believed, data is so much more than information, it’s potential. Potential quantified. I became fascinated by how machine learning could be leveraged beyond the lab, and that curiosity ultimately led to the creation of Camouflet. Camouflet is fundamentally about intelligent decision-making. We built the platform to empower businesses with adaptive, data-driven pricing strategies that keep them competitive in a chaotic market. By integrating predictive analytics, demand forecasting, and scenario testing. It’s a different industry from where I started, but the mission remains the same, using technology to make sense of complexity and create better outcomes.
None of us can achieve success without some help along the way. Is there a particular person who you are grateful for, who helped get you to where you are? Can you share a story?
I’ve been fortunate to have incredible mentors, both in academia and industry. My doctoral advisor was a force of nature, tough, brilliant, and relentless in pushing me to succeed in ways few people have. But when I think about the person who truly made the biggest impact on my journey, it’s actually my partner from my time as a doctoral student, Jack.
We were both broke, overworked graduate students, he was studying fine art at Goldsmiths, and I was deep in molecular biology at Imperial. In a lot of ways, we couldn’t have been more different, but that contrast made our life together something special. The merging of our lives as scientist and artist, in life and love, was an exceptional thing. He had this way of making even the smallest moments feel magical. I’d wake up to find a sketch he had drawn of me sleeping on my bedside table. He would plan these small but thoughtful escapes from the stress, trips, dinners, anything to carve out space where science and deadlines didn’t dominate. That balance was everything.
I think everyone deserves a Jack, someone who reminds you that your work isn’t your entire identity, that success isn’t just about accomplishments, and that even in the most high-pressure environments, life still has room for beauty.
You are a successful business leader. Which three character traits do you think were most instrumental to your success? Can you please share a story or example for each?
I think success is rarely about just one thing, it’s a mix of mindset, adaptability, and the people you surround yourself with. But if I had to narrow it down, the three traits that have been most instrumental for me are intellectual curiosity, resilience, and decisiveness.
I never set out to be in business. My background is in molecular biology, and my academic work started in public health research. But what has always driven me is the need to understand how things work. I’m exhaustingly curious, and I don’t accept surface-level answers, I want to take things apart, find patterns, and build something better. That mindset is what led me from studying disease models at NYU to creating the first computable phenotype for Long COVID, and ultimately, to founding Camouflet. The ability to see connections between different disciplines is what has allowed me to develop technology that applies to real-world problems in unexpected ways.
Startups are hard. There are moments when things don’t go as planned, when funding is precarious, when people doubt your vision. Early on, when I was building Camouflet, there were multiple moments where it would have been easier to pivot into something more conventional. Investors always want to know why you’re different, but they also fear what they don’t recognize. It took a lot of resilience to keep pushing forward and prove that a closed-loop AI system was not just possible but necessary against an over-reliance on third-party tools.
At a certain point, overanalyzing becomes paralysis. One of the first lessons I learned was that you can’t wait for perfect information and the need to exemplify decisiveness. At Camouflet, we built everything in-house, but making that call early on meant rejecting the industry-standard approach of outsourcing. That was a high-risk, high-reward decision, but it’s the reason our platform is now seen as a category-defining solution rather than just another optimization tool.
Thank you for all that. Let’s now turn to the main focus of our discussion about how AI leaders are keeping AI safe and responsible. To begin, can you list three things that most excite you about the current state of the AI industry?
AI is a tool that’s neither inherently good nor bad. It’s a reflection of the people building it. Its safety depends on who is training the models, what data is being used, and what incentives are at play. While there are legitimate concerns about its misuse, there’s also a lot to be excited about. The three things that stand out to me most right now are the shift toward proprietary AI models, advances in AI interpretability, and the increasing demand for accountability in AI-driven decision-making. I see three major shifts that are actively improving the landscape: companies investing in proprietary AI instead of relying on external APIs, new breakthroughs in AI interpretability that allow for real auditing, and growing pressure, from both regulators and consumers, for accountability when AI makes critical mistakes.
Conversely, can you tell us three things that most concern you about the industry? What must be done to alleviate those concerns?
My biggest concerns about AI aren’t about the technology itself but how it’s being developed and controlled. First, too much power is concentrated in the hands of a few companies, creating an unchecked monopoly over information and decision-making. That kind of control stifles competition and limits transparency. We need policies that ensure decentralization, open standards, and real competition in AI development.
Second, AI-generated misinformation is scaling faster than our ability to detect or regulate it. Deepfakes, fabricated news, etc., are already blurring the line between reality and manipulation. We need stronger verification systems, traceability tools, and policies that hold platforms accountable for AI-driven deception. Otherwise, trust in digital information will erode entirely.
Finally, regulation is either nonexistent or too reactionary. Governments either lag behind the technology or overcorrect with policies that entrench existing players while crushing startups and independent research. Instead of vague ethical guidelines or blanket bans, we need clear, enforceable AI standards that prioritize transparency, without stifling innovation. AI should be a tool for progress, not a weapon for misinformation or a tool of monopolistic control.
As a CEO leading an AI-driven organization, how do you embed ethical principles into your company’s overall vision and long-term strategy? What specific executive-level decisions have you made to ensure your company stays ahead in developing safe, transparent, and responsible AI technologies?
AI’s biggest risks aren’t technical, they’re structural. Power is becoming dangerously concentrated in the hands of a few major players, with companies like OpenAI, Google, and Microsoft controlling the models that shape everything from information access to hiring decisions. We have antitrust protections to prevent monopolies in commerce, yet we’re watching an unchecked monopoly on information itself take hold. A monopoly that dictates not just what people see, but how they think
At the same time, we’re seeing an explosion of AI-generated misinformation that spreads faster than fact-checking can keep up, eroding public trust in everything from news to scientific research. And while regulation is necessary, most governments either don’t understand AI well enough to craft effective policies or overcorrect in ways that stifle innovation. The solution isn’t about stopping AI, it’s about decentralizing control, enforcing transparency, and building regulatory frameworks that hold companies accountable without crushing progress. Companies need to take responsibility for the AI they deploy, and regulators need to focus on oversight that ensures safety and fairness without handing even more power to the incumbents.
Have you ever faced a challenging ethical dilemma related to AI development or deployment? How did you navigate the situation while balancing business goals and ethical responsibility?
One of the toughest ethical dilemmas we faced was deciding how much autonomy our AI should have in pricing decisions. A potential client wanted fully automated, profit-maximizing price adjustments with no human oversight. When we ran simulations, the AI began charging certain customers significantly more based purely on behavioral data, not true market conditions. The model wasn’t intentionally unethical, but just because AI can do something doesn’t mean it should. We made the call to require human oversight in all AI-driven pricing recommendations and build transparency tools that explain why a price was suggested. Some argued full automation would drive higher profits, but trust and transparency matter more than short-term gains. The companies that optimize without oversight will eventually face regulatory and reputational fallout, it’s just a matter of when.
Many people are worried about the potential for AI to harm humans. What must be done to ensure that AI stays safe?
AI itself isn’t dangerous. It’s the way people design, deploy, and regulate it that determines whether it helps or harms. The biggest risks come from lack of oversight, blind trust in black-box models, and the rush to deploy AI at scale without accountability. The solution isn’t to slow AI down, but to demand transparency, enforce human oversight in high-stakes decisions, and hold companies accountable for failures. No AI system should be making decisions that impact people’s lives, whether in hiring, healthcare, or criminal justice, without a clear, auditable explanation of how it arrived at that outcome. Regulators need to move beyond vague ethical guidelines and implement clear, enforceable standards for explainability, bias detection, and accountability. AI will only be as safe as the people controlling it, so the real challenge isn’t stopping AI from harming humans, it’s stopping humans from using AI irresponsibly.
Despite huge advances, AIs still confidently hallucinate, giving incorrect answers. In addition, AIs will produce incorrect results if they are trained on untrue or biased information. What can be done to ensure that AI produces accurate and transparent results?
AI hallucinations and misinformation are direct results of garbage in, garbage out. If a model is trained on flawed, biased, or outright false data, it will confidently produce bad results. The solution isn’t just better algorithms, it’s better data curation, stronger verification systems, and full transparency in AI decision-making. Companies should be auditing training data, filtering out misinformation, and ensuring models don’t blindly reinforce biases. But even with perfect data, AI will still make mistakes, which is why we need built-in mechanisms for error detection, user verification, and clear disclaimers when confidence levels are low. At Camouflet, we’ve made explainability a core principle, no AI-driven decision is deployed without human-auditable reasoning behind it. If a model can’t justify its output, it has no business being in production.
Here is the primary question of our discussion. Based on your experience and success, what are your “Five Things Needed to Keep AI Safe, Ethical, Responsible, and True”? Please share a story or an example for each.
Keeping AI safe, ethical, responsible, and true isn’t about following abstract principles — it’s about building AI systems with real guardrails that prevent harm before it happens. There are five key things every AI company should be doing:
- Full Data Transparency: AI is only as good as the data it’s trained on. If companies don’t disclose where their training data comes from, they can’t be trusted. We’ve seen disasters like AI-powered hiring tools that discriminated against women simply because they were trained on decades of biased hiring data. At Camouflet, we audit every dataset we use and reject black-box training sources. No transparency, no trust.
- Human Oversight in High-Stakes Decisions: AI should never operate unchecked in critical areas like healthcare, finance, or criminal justice. The UK’s Post Office scandal, where an AI-driven accounting system falsely accused hundreds of employees of fraud, proved how dangerous blind trust in automation can be. Every AI-driven pricing decision we make at Camouflet has a human-in-the-loop system to prevent errors and unintended consequences.
- Explainability and Justification: If an AI system can’t explain its decisions, it shouldn’t be making them. Look at Australia’s RoboDebt disaster, where an automated system falsely calculated welfare overpayments, devastating thousands of people financially. It took years to unravel because no one could explain how the AI was making its calculations. At Camouflet, every pricing recommendation must have a clear, auditable reasoning chain that can be manually reviewed.
- Accountability for Failures: AI companies love to take credit when things go right but blame “the algorithm” when things go wrong. That has to change. When self-driving Teslas caused accidents, the company initially refused to take full responsibility, saying drivers should have been paying attention. AI needs clear liability structures, if an AI decision harms someone, the people who built and deployed it should be accountable.
- Regulation That Makes Sense: We need laws that ensure AI is safe without stifling innovation. Europe’s AI Act is moving in the right direction, requiring transparency and human oversight for high-risk AI, but the U.S. still lacks meaningful AI regulations. The danger isn’t just underregulation, it’s bad regulation that hands more power to existing tech monopolies. AI governance should focus on transparency, fairness, and accountability without reinforcing existing corporate dominance.
AI safety is about ensuring progress doesn’t come at the expense of people’s rights, livelihoods, or safety. Companies that ignore this will face regulatory, legal, and public backlash. The ones that take it seriously will define the future.
Looking ahead, what changes do you hope to see in industry-wide AI governance over the next decade?
AI governance needs enforceable standards, not just guidelines. Over the next decade, I want to see mandatory transparency in high-stakes AI decisions, real accountability when AI causes harm, and policies that prevent tech monopolies from controlling AI development. No system should be deployed without an auditable explanation of its decisions. Companies must be liable for AI failures, not able to hide behind “the algorithm.” And just as we have antitrust laws for commerce, we need protections against a handful of corporations monopolizing AI. Done right, AI can drive progress, left unchecked, it risks consolidating power in ways we can’t undo.
What do you think will be the biggest challenge for AI over the next decade, and how should the industry prepare?
The biggest challenge for AI over the next decade will be trust. Public skepticism is growing as AI becomes more powerful but also more prone to bias and misinformation. If people don’t trust AI, they won’t adopt it, or worse, they’ll resist it entirely. The industry needs to prepare by prioritizing transparency and decentralizing control. AI systems must be explainable, with clear reasoning behind decisions. Companies must take responsibility for AI failures instead of blaming “the algorithm.” And we need safeguards against a handful of corporations monopolizing AI development. Without these changes, AI won’t just face regulatory pushback, it will lose the public’s confidence, which is much harder to regain.
You are a person of great influence. If you could inspire a movement that would bring the most good to the most people, what would that be? You never know what your idea can trigger. 🙂
AI should be a tool for progress, not a mechanism for reinforcing inequality. That starts with who gets to build it. Too often, AI reinforces existing biases because the people designing it come from the same narrow backgrounds. At Camouflet, we don’t just talk about diversity, we invest in it, whether through our Diversity in Tech Fund contributions, mentorship programs, or ensuring inclusive hiring practices within our own company. The reality is, AI built by a homogenous group will always have blind spots. If we want AI that serves everyone, we need people from all backgrounds shaping its future. This isn’t just an ethical issue, it’s a practical one. The more perspectives we bring into AI development, the better, fairer, and more effective the technology will be. That’s the movement I want to drive, one that ensures AI is built by, and for, the full spectrum of human experience.
How can our readers follow your work online?
For my professional work with Camouflet, visit https://www.camouflet.co to learn more about our approach to AI-driven pricing and innovation. I’m also a published author, and you can find more about my writing at https://www.jeffradwell.com or on my Amazon author page. I love to discuss, debate, and opine, and there are numerous ways to connect with me through my website. I welcome anyone interested in AI, technology, or writing to reach out, I’m always eager to share perspectives and continue the conversation.
Thank you so much for joining us. This was very inspirational.
I appreciate the thoughtful questions and the chance to share my perspective. AI is shaping the future, and it’s up to us to ensure that future is ethical. These conversations matter and the more we discuss the impact of AI openly, the better equipped we are to build something responsible and lasting. Thanks again for having me.
Guardians of AI: Jeff Radwell of Camouflet On How AI Leaders Are Keeping AI Safe, Ethical… was originally published in Authority Magazine on Medium, where people are continuing the conversation by highlighting and responding to this story.