How LLMs Are Being Used in Scams

Every new technological development promises a host of benefits. Unfortunately, as with any tool, it can also be used to harm as much as help. This is true of AI, and Large Language Models (LLMs) in particular. In this article, we’re looking at LLM scams and how AI can be used to prey on people.

Enhancing Authenticity

Much as you might use an LLM to enhance your words or improve your writing, so too do scammers use AI to generate believable and convincing text that you might accept as true, especially if your first language isn’t the one it uses. The authoritative spelling, grammar, and knowledge improve the illusion and give off the impression of authenticity. Let’s see some of the ways this is utilized.

Scam Detectors Most Trusted Websites in Online Security

Guard.io (100): Protect your digital world on any device – Guardio stops scams and phishing in their tracks.
Incogni.com (100): Delete your personal data from the internet and protect against scams and identity theft.
ExpressVPN (100) Stay secure and anonymous online - Best VPN Out There

LLM Scams

There’s no definitive list of AI scams, as the technology can be used in many supportive or primary roles. We do have some major examples to share, though:

Jailbreaking

Many LLMs have guardrails in place to prevent abuse by users. That said, these can be circumvented in a process called jailbreaking. This can be done by posing hypothetical scenarios or posing as a different person.

Some LLMs have no guardrails and can be used for any end. This enables scammers to move past restrictions and attack you.

Crypto Scams

Crypto is a high-risk, high-reward form of investment, popular with so many scammers that you don’t know who to trust. So if a major figure like Michael Saylor of MicroStrategy offered you a surefire road to success on a secret channel, you might believe him. But deepfake scams use likenesses of CEO figureheads such as these to guide users to fake sites or set up fake overlays. Once on these sites, the victim is easy prey for their money to be stolen.

Impersonation Scam

Beyond crypto, scammers can also use LLMs to commit other types of fraud, like phishing, romance scams, and finance scams. AI-generated deception is no longer limited to text but also includes convincing videos that resemble conference calls, making everything from investment opportunities to job interviews a chance for some poor victim to be scammed out of their money or personal information.

Depending on the technology, these scams can be dragged out for some time, like pig butchering scams, where scammers tap into victims’ emotions to gain access to their accounts.

SEO and GEO Ranking

Keywords rule the major search engines. If you want to feature in a search, you need to know you’re using the right keywords. Likewise, scammers use LLMs to help them find the right keywords for the right audience, to improve their SEO rankings and be closer to the top of searches, and thus more likely to get clicks.

In addition, they’re also taking advantage of Generative Engine Optimization (GEO) to have their content scraped and used in chatbot searches. Victims get information from their favorite chatbot, only to be directed to untrustworthy sources where scammers can gain access to key information.

Task Automation

You love LLMs for automating repetitive tasks. Scammers do too. The technology involved in these bots can make scraping personal information, gathering data of potential victims, responding automatically to callbacks, and other tasks a breeze. This enables larger-scale operations with fewer resources, and is quite worrying.

Verify Another website

Are you just about to make a purchase online? See if the website is legit with our validator:

vldtr^®