Hearing Isn't Believing Anymore.

A few years ago, AI voice cloning sounded like science fiction, but today, it's a real business risk.

In this episode of Glitchd, I explore how AI-generated voices and deepfakes are changing the way organisations think about trust, fraud, and verification.

Hearing someone's voice is no longer proof it's really them.

As AI tools become more powerful and accessible, fraudsters are increasingly using voice cloning and deepfake technology to impersonate executives, colleagues, suppliers, and trusted contacts.

In this episode, I discuss:

• How AI voice cloning works
• Why familiar voices are no longer enough to verify identity
• The growing threat of AI-enabled fraud
• Why traditional trust signals are becoming less reliable
• The governance and risk implications for organisations
• Practical steps businesses can take today to reduce their exposure

The organisations that adapt now will be far better positioned than those that wait until after an incident occurs.

Whether you're a business owner, executive, finance professional, risk manager, compliance lead, or simply interested in the real-world impact of AI, this is a conversation worth having.

🎙️ Glitchd is hosted by Sharon, AI Governance Professional, Founder of SKS Professional Services, helping organisations navigate AI governance, risk, compliance, and responsible AI adoption.

https://skspsl.com/

Transcript

SPEAKER_00 0:06

I'm Sharon and I work in AI governance and I'm the founder of SKS Professional Services where I help businesses adopt AI responsibly, manage risk, and put the right governance in place so they can innovate with confidence. This is Glitched. New episodes every Tuesday, let's get into it. I want to start with a simple question. If your CEO called you right now and asked you to urgently approve a payment, what would you do? Most people listening are probably thinking, well, I'd check first if I was in that department, and that's fair. But here's the thing. The people who have fallen victim to AI voice scams thought exactly the same thing. The reason I wanted to talk about this today is because I think a lot of businesses globally are still treating this as something that is coming. But I don't think it's coming anymore. I think it's already here. A few years ago, if someone told you that AI could convincingly copy a person's voice from a short clip of audio, it sounded like science fiction. But now it's a subscription service. And trust me, I've tried one of these services. And if you've ever appeared on a podcast, spoken at an event, uploaded a video to LinkedIn or done a webinar, etc., there's a good chance enough audio of your voice already exists online. Now let's talk about the real issue here. And this is not really technology related, it's a trust story. For years we've relied on certain signals. For example, you recognize someone's voice, you recognize their face, you recognize the way they speak, and you don't really think about it, right? Your brain does that work for you. So if your colleague calls you, your boss calls you, you know who it is calling you. The problem is that this assumption is becoming less and less reliable. We're now seeing more and more examples of fraudsters using AI-generated voices and, in some cases, AI-generated video to impersonate trusted people. There have been some major, major cases of this happening this year alone. Think of your senior leaders, your colleague that sits next to you every day, business partners, the people you already know, and this is what makes it dangerous. And this isn't really about being careless, it's just that we are human, right? We like to trust. So if someone says to you, can you help me with something urgently? a lot of people aren't immediately thinking, oh, this could be an AI-generated voice clone, let me check it out further, let me investigate. They're probably thinking, How quickly can I get this sorted? I mean, you've already got so many things to do in your day. So when your colleague who's working from home calls you and says, I need help urgently, hmm, how likely is it that you're just going to take some action? So, why should businesses care about this? You might be considering, well, I'm sure this only happens to massive companies with a lot more to lose or to give. Not really. Frauders tend to go where the opportunity is, and the barrier to entry is dropping all of the time. So what used to require specially skills and expensive tools can now be done with a software that's increasingly accessible and affordable. It's not just that the technology has improved, it's actually that it's becoming easier to use. And if we've been practical about it, most businesses have spent years investing in technology controls. Think of your monitoring, access management, firewalls, security software. All of these things are important. But a lot of decision making still comes down to trust, a familiar voice, a familiar face, a request from someone senior, from your boss. And I think this is where many organizations remain exposed. So where does the governance come in? And this is one of the reasons I'm always banging on about governance. And I know governance isn't the most exciting word in the world, but good governance isn't just policies and paperwork. You see, at its heart, it's about helping organizations make sound decisions when the environment changes. And the environment is changing very, very quickly with AI. The question isn't whether AI can clone a voice, because it can. The question is what that means for your processes. Because if your approval process relies on recognising someone's voice, you have a vulnerability. If your finance team, for example, feels unable to challenge a request from a senior executive, you've got a vulnerability. If a payment can be authorized based on a phone call alone, you have got a vulnerability. Do you see the trend here? So the risk really isn't just about the AI. The risk is continuing to operate as though nothing has changed. Right, what should businesses do? The good news is that this isn't going to be solved by buying another tool. You don't need to increase your budget, you don't need to increase your spend, because most of the protection, as I've mentioned, will come from processes, evaluating your processes. For example, if someone requests a payment over the phone or during a video call, there should be a separate verification step before any money moves. Simple. The same applies to a supplier bank detail changes, payroll requests, sensitive information requests, anything that involves money and or significant business decisions. So I'd encourage businesses to have these conversations right now, as soon as possible, not after an incident occurs. Sit down and ask yourselves, would our current process stop a convincing voice clone? Would our employees know what to do in that situation? Would they feel comfortable saying, I'm sorry, but I need to verify this? Even when they think they are speaking to the CEO? Because if the answer is no, then you've already identified something that's worth fixing. Looking ahead. For decades now, hearing someone's voice meant something. It was enough. Seeing someone's face on a screen was enough. Those things still matter, but they're not enough on their own anymore. And I don't think enough businesses are fully caught up with that reality. The businesses that will navigate this world are not going to be the ones spending the most money. When it comes to AI and staying safe, making responsible decisions and doing this profitably, I've always said and I will continue to say, it's not about who has the biggest budget or who has the biggest spend. These successful businesses will be the ones that are able to recognize one, what's changed, and two, to adapt accordingly. I'll finish off with an example that I have experienced myself. I was speaking with my tax advisor on WhatsApp. We chat all the time, and I sent over some instructions by texting, and then I followed up to clarify with a voice note. There was no reply, put my phone away, was doing some work, and I returned to my phone and found there was a missed call. And at that time I wasn't able to call back, so I sent another text with a brief voice note to say, sorry, I'm really busy at the moment, but what is it you wanted to chat about? And they just text back and said, uh please just give me a call and then I can explain further. After I called, my tax advisor advised me that they wanted to make sure they were speaking to me, to Sharon, because they are aware of all of these scams and these fraudsters floating about. So even though this was our chat on WhatsApp and she knows my voice, she wanted to check. She aired on the side of caution. I'm Sharon. I work in AI governance and I'm the founder of SKS Professional Services, where I help businesses navigate AI adoption, governance, risk, and compliance in a way that supports innovation while protecting trust. I'm all about scaling responsibly and profitably. If you found today's episode useful, send it to someone in your business, in finance, operations, risk, send it to a friend. Thank you for listening, and I will see you next Tuesday.