Deepfakes, cloned voices and fake investments - how AI is changing fraud

AI is lowering the cost of deception

Tools that once required specialist teams are now easy to access. Voice cloning, image generation and persuasive scripts can be assembled quickly and cheaply, which means more scammers can run more convincing campaigns.

The attack is still social engineering at its core. AI simply makes the emotional hook stronger and the fake content more believable.

What deepfakes look like in everyday life

For older adults the risk is rarely a cinematic fake video. It is more often a believable voice note, a short urgent call, a fake celebrity investment ad or a message that appears to match a familiar tone.

The goal is always the same: create trust fast, create urgency faster, and push the victim toward a transfer, code or click.

Fake investment journeys are becoming smoother

AI is particularly useful for scammers running fake investment schemes. It helps them write persuasive ads, personalise follow-up messages and create websites that look polished and current.

That matters because many victims do not lose money in the first minute. They are guided through a process that feels increasingly professional at every step.

How to defend yourself

Do not trust realism on its own. A realistic voice is still only a voice. A polished website is still only a website. Verification has to come from outside the message: a known phone number, an official bank site typed manually, a second person you trust.

In ODYSSEY we treat AI scams as a behaviour problem, not just a technology problem. The best defence is still a trained pause, a short checklist and a verification habit.