Fire with fire: AI-powered security outlook from Derek Swift

FNG Exclusive Interview… With the rise of digital platforms and peer-driven financial tools, the threats targeting online systems have evolved too. We sat down with Derek Swift, Head of Client Funding Facilities at leading online broker Deriv, to talk about the state of online fraud, how AI is changing the game, and how Deriv is building more resilient systems, especially within peer-to-peer features.

FNG: Hi Derek, and thanks for joining us today. The online space has become a hub for quick and convenient financial activity, but it also feels riskier than ever. Is there a spike in digital scams, or is it just that we’re doing more things online than before?

Derek: The digital space is definitely seeing a spike in scam attempts. There are now smarter, more calculated attacks. It’s no longer about the mass spam emails, as digital literacy has kind of limited their effectiveness. Instead, a wave of AI-generated phishing, personalised scam attempts, and even voice cloning are on the rise. Fraudsters are evolving as fast as the platforms they target.

In spaces where people interact directly, there’s naturally more room for manipulation. That’s why when we built our payment solution we didn’t just focus on usability. We were asking with every step of the way, “How can this be secured?”

FNG: So where are people most at risk? Is it still just bad passwords and dodgy links or more efforts are needed to stay safe online?

Derek: Those are definitely still part of it, but the real danger is mostly behavioural. Scammers nowadays are not targeting people’s tech, but their trust, which is proving to be quite a weak link. We’re talking about social engineering, through which they trick users into sharing sensitive info, clicking a fake link, or moving outside of a secure environment.

For example, someone might pretend to be a verified user and ask to continue a conversation off-platform. As other systems may not be as secure as ours, we made it mandatory for all peer-to-peer conversations to take place within the Deriv platform. If someone’s trying to take things elsewhere, that’s a bad sign that we’ve trained the system to recognise and flag immediately.

We’re also seeing account takeovers, phishing emails that look official, and increasing attacks targeting mobile devices, especially in regions where these devices are the primary connection point to the internet for many people.

FNG: What does proactive security look like today?

Derek: Generally speaking, it’s about creating layers of trust. Think multi-factor authentication, encrypted activity, alerts for strange behaviour, and making sure that users understand what to look out for.

When building Deriv’s payment features, we implemented a secure process that involves multiple steps. Users need to confirm each stage of the transfer, and nothing is finalised without all sides of the transaction doing their due diligence.

We’ve also integrated automated monitoring tools that can spot suspicious activity patterns early. But tools are only half the story. Users’ awareness is essential. We can enable that by a clear interface design, confirmation screens that slow people down, and continuous awareness efforts along the way.

FNG: AI is a hot topic right now. How is it shaping the digital safety landscape?

Derek: AI’s influence is massive on both ends. On our side, machine learning helps us detect behavioural anomalies, like someone logging in at an odd hour or trying to interact with lots of users in quick succession. That’s something a human might miss, but an AI model can spot instantly.

On the other hand, fraudsters are using AI to mimic writing styles, generate believable emails, and even simulate real-time conversations. Our system has flagged scam attempts that look nearly indistinguishable from genuine support messages.

That’s why security has to evolve constantly. You can’t set it and forget it. We retrain our models regularly, using anonymous behavioural data to improve detection. But again, AI and tech in general are only part of the solution. The rest is creating and securing safe environments where users know what to expect and can act confidently.

FNG: You’ve mentioned trust a few times. In practical terms, how do platforms earn that trust?

Derek: Transparency, predictability, and clear communication. Trust isn’t about making big promises, but rather following through on small things consistently.

In Deriv’s case, we set out to make users’ interactions with us and each other safer by default. We don’t allow off-platform communications because they remove a layer of protection. And if something goes wrong, our support team is trained to act fast, with full visibility of the interaction.

It’s also about not overwhelming users with jargon or flashy warnings. People feel safer when the system is intuitive and guides them gently but clearly. Trust comes from knowing you’re in control, not just hoping you are.

FNG: Looking ahead, what can we expect in the next wave of cybersecurity and fraud prevention?

Derek: We’re going to see more personalisation in both attacks and defences. Scammers will get better at mimicking real people, and security systems will get better at understanding individual behaviour patterns.

Regulations are also likely to evolve, especially around identity verification and data handling. And platforms will have to strike a balance between robust security and smooth user experience.

One thing is clear, though: no single system or company can tackle this alone. It’s going to take collaboration across industries, regions, and even between competitors to stay ahead. We all have a role to play in making these systems safer, especially as they become more peer-driven and global.