Transcript
Manoj Tandon:
Hello everyone, and welcome to another episode of Dark Rhiino Security’s Security Confidential. I’m your host, Manoj Tandon. Before we get started, please remember to hit the like and subscribe button. It helps us continue bringing you this content and allows us to feature these great guests.
Today, I’m honored to welcome Dr. Rizwan Sheikh to the program. Rizwan is an old friend of mine—we were college roommates back in the 1980s at Ohio State University, long before AI and cloud computing were part of everyday conversations. Today, Rizwan is the CEO of Global AI Excellence, a company he founded, and the inventor of a groundbreaking AI governance model. He’s one of the leading voices helping organizations deploy AI responsibly. He has an extensive consulting background with Deloitte and Fortune 500 companies, and he teaches AI strategy at institutions like Harvard and MIT. Rizwan, welcome to the show.
Dr. Rizwan Sheikh:
Thank you, Manoj, for the very generous introduction. I’m doing well, and yes—it’s been a long road since Ohio State and Morrill Tower.
Manoj Tandon:
It really has. For those who don’t know, Morrill Tower was… an experience. We’ll leave those stories in the past. Let’s start with your background beyond OSU. Tell our audience a bit about what led you here.
Dr. Rizwan Sheikh:
Let me start with some recent statistics that frame why this conversation matters. According to a 2025 TrustMarkets survey, 93% of organizations are using AI in some form, but only 7% have embedded AI governance into their deployment pipelines. An IBM report from 2025 shows that the average AI-related security breach costs $5.7 million per incident. And only 27% of corporate boards have formally added AI governance to their charters.
These gaps have real consequences. Deloitte recently had to refund nearly $300,000 to the Australian government because a report they submitted—generated by AI—contained fake citations. The financial cost was one thing, but the reputational damage was far greater. These are not hypothetical risks. They’re happening right now.
Manoj Tandon:
So why governance? Why don’t existing governance models work?
Dr. Rizwan Sheikh:
That’s a great question. I’m a strong believer in AI innovation. AI can do incredible things if proper governance and guardrails are in place. My journey into AI governance started about four years ago, when tools like ChatGPT began emerging. Like many others, I was excited.
One night—around 2:42 a.m.—I was in my garage in Dallas experimenting with an AI model. I asked it to generate a five-page report with citations. In under a minute, it produced a well-written, polished document. The language was flawless. But when I checked the references, every single citation was fake. All of them.
That’s when it hit me: AI hallucinates, and it hallucinates with confidence. That’s a massive risk—not just for individuals, but for organizations. That moment is when my AI governance model was born.
Manoj Tandon:
And hallucinations aren’t going away anytime soon.
Dr. Rizwan Sheikh:
Exactly. It’s inherent in how these systems are trained. AI will always give you an answer—it doesn’t say, “I don’t know.” And it presents false information in a way that makes it seem credible. That’s dangerous.
We’ve seen real-world consequences: a California attorney fined $10,000 for submitting fake AI-generated case citations, and tens of thousands of families in Europe falsely accused of fraud due to AI-driven profiling errors. In some cases, children were taken from families. These are devastating outcomes.
Manoj Tandon:
Is this user error? Should people be prompting AI to validate itself?
Dr. Rizwan Sheikh:
Not necessarily. When users ask for “proper citations,” they reasonably expect those citations to be real. While prompting can help in some cases, hallucination is fundamentally a system-level problem. Garbage in, garbage out still applies.
Manoj Tandon:
So how do you govern something that people use because it’s convenient?
Dr. Rizwan Sheikh:
That’s exactly the challenge. Convenience always wins—unless governance is built in from the start. My AI governance model is patent-pending and combines advanced AI, human-in-the-loop decision-making, and blockchain-based audit trails. It includes a citation verification engine and mandatory human review for high-risk use cases, such as healthcare, finance, and government services.
Manoj Tandon:
Healthcare seems like an obvious place where AI needs oversight.
Dr. Rizwan Sheikh:
Absolutely. False positives and false negatives in healthcare can be fatal. Our framework allows AI to do preliminary analysis—such as reviewing MRIs—but a licensed physician must validate the results before anything is shared with a patient. AI augments expertise; it does not replace it.
Manoj Tandon:
Some would argue that slows things down.
Dr. Rizwan Sheikh:
Only slightly—and only where it matters. We use a risk-based approach: low-risk cases can be automated, but high-risk decisions require human oversight. The productivity cost might be 5–10%, but the benefit is preventing catastrophic outcomes. That tradeoff is worth it.
Manoj Tandon:
What about non-healthcare use cases, like logistics?
Dr. Rizwan Sheikh:
The same principle applies. Internal forecasting may be low risk, but customer-facing or compliance-sensitive decisions require oversight. AI hallucinating in logistics could mean missing entire sales seasons. Governance ensures AI supports business goals rather than undermining them.
Manoj Tandon:
So where do you start with a client?
Dr. Rizwan Sheikh:
We begin with an AI Governance, Risk, and Compliance assessment—AI GRC. We’ve built one of the largest AI GRC question databases in the industry, covering more than 12 sectors, including healthcare, finance, energy, government, and manufacturing.
From there, we assess maturity, identify risk posture, map gaps against regulations and internal policies, and provide a phased remediation plan—30, 60, 90, 180, and 365 days—along with ROI analysis.
Manoj Tandon:
How long does that process take?
Dr. Rizwan Sheikh:
Traditional assessments take 12–15 weeks. We can deliver a preliminary assessment in about two weeks—an 80% reduction—by leveraging AI responsibly.
Manoj Tandon:
Who is this best suited for?
Dr. Rizwan Sheikh:
Organizations with revenues above $500 million benefit most, but we also support mid-sized enterprises with flexible pricing models.
Manoj Tandon:
You mentioned regulations earlier. Can you explain the EU AI Act?
Dr. Rizwan Sheikh:
Any company doing business in the EU falls under the EU AI Act. Penalties can reach €35 million or 7% of global annual revenue per incident. A biased loan decision or flawed healthcare model could trigger massive fines. Similar laws are emerging in the U.S., including the Colorado AI Act and NIST AI Risk Management Framework guidance.
Manoj Tandon:
Where is all of this heading?
Dr. Rizwan Sheikh:
By 2030, AI investment will reach between $800 billion and $2 trillion. AI governance alone is projected to be an $8.5 billion market. The winners won’t be companies that move fastest—they’ll be the ones that deploy AI responsibly with strong governance foundations.
Manoj Tandon:
That’s a powerful takeaway. Before we wrap up, what would you like our audience to know?
Dr. Rizwan Sheikh:
I publish weekly content on LinkedIn and speak regularly on AI governance, risk, and compliance. I encourage your audience to follow me and Global AI Excellence on LinkedIn. If you want to build or use AI, don’t be afraid of it—but don’t deploy it without guardrails.
Manoj Tandon:
Rizwan, this has been an outstanding conversation. Thank you for joining us, and please keep us updated as Global AI Excellence grows.
Dr. Rizwan Sheikh:
Thank you, Manoj. It was a pleasure, and I look forward to coming back.
Manoj Tandon:
Take care.
Chapter Titles:
00:00 Intro
02:34 Our Guest
03:33 93% of Employees are using AI
05:04 Why don’t existing governance models work?
06:15 ChatGPT Sources are fake
07:17 A.I Hallucinates
12:06 How do you govern Ai?
22:02 When should you govern Ai?
31:47 What should AI look like for my company?
33:45 European Union AI Act
39:05 Believing False AI statements
43:20 How can you build governance to something that changes daily?
45:43 How do I get into AI?
48:06 Connecting with Dr. Rizwan