Fighting AI fraud with AI: that’s this startup’s mission
Whether it’s a Nigerian prince asking for an advance or a threatening message from the IRS about overdue taxes, we’ve probably all been targeted by scammers at some point.
It’s big business for criminals: Despite increased awareness, Americans $10 billion lost to this type of fraud in 2023 alone.
The problem is getting worse for a number of reasons, including a relative newcomer: generative AI. The same technology we use to polish our resumes, generate realistic images, and write witty poetry can be used by scammers to rip us off. With the rise of popular chatbots like OpenAI’s ChatGPT and Google’s Gemini, scammers have their own superhero sidekicks to craft believable emails and text messages asking for money.
“We like to say the bad guys have started using AI and the good guys have started using education,” said Scamnetic CEO Al Pascual. “Your bank tells you, ‘We will never do this. We will never do that. Never click on this.’”
While financial services firms are using AI internally to identify and prevent scams, we consumers haven’t had many options to use AI to protect ourselves. But that’s starting to change.
When Pascual was senior principal and enterprise solutions lead at credit reporting agency TransUnion in 2023, he noticed that scams were getting worse.
“There was something that happened with COVID where the criminals had an epiphany … especially after the whole benefit fraud thing went down … it seemed like they realized it’s just easier to ask a bunch of people for money because eventually someone will say yes,” he said. “I realized the opportunity was pretty significant.”
TransUnion is not in the business of fighting scams, so Pascual stepped out and founded Scamnetic in September 2023.
KnowScam and IDeveryone
The startup focuses on two main areas.
The first is inbound communications such as email, text messages, and social media posts. Scamnetic analyzes these messages to determine if they are legitimate or malicious using a product called KnowScam.
The model uses machine learning and natural language processing to evaluate content based on what it learned during training, which includes “thousands of good posts and thousands of bad posts,” Pascual said.
(Machine learning is a branch of AI that uses data and algorithms to mimic the way we humans learn. Natural language processing uses machine learning to help machines understand and communicate with human language.)
KnowScam scores messages on a scale from one, which is low risk, to four, which is high risk. This score is based on message details such as sender, images, links, and attachments and whether any of these components can be matched to patterns for previous scams.
Scamnetic can also help you manage risks when reaching out to people you don’t know, such as on dating sites or e-commerce platforms like Facebook Marketplace. For this, the company has a product called IDeveryone.
For example, before you send money to a stranger with a product for sale, you can ask them to verify themselves with this tool. After you add their phone number or email address, the person will receive a message asking for their name, location, and permission.
“This is a consent process because we go to non-public data sources and background verify you,” Pascual said.
Tampa, Florida-based Scamnetic plans to launch a direct-to-consumer subscription service in the fourth quarter. It is currently working with partners including banks, credit unions, telecom companies and insurance companies, which offer KnowScam and IDeveryone to their customers as a free benefit or as part of a premium package.
Deepfakes
In the future, Pascual wants to expand the channels Scamentic covers, with more activity on social media. He also plans to launch video communications to help users detect deepfakes, or images and videos that have been altered to show something that never actually happened.
“That’s a big thing on our roadmap for the relatively near term, because that’s an impossible problem for consumers,” he said of deepfakes. “It’s one thing to look at an email and say, ‘Does that look like a good PayPal logo?’ But when you see someone on screen in a video — and some of the high-quality videos are really good — it can be almost impossible for someone to know that.”
Scamnetic raised $1.35 million in a pre-seed round in March.
Investors include early stage investment firms Ruxton Ventures, Roo Capital and SaaS Ventures, as well as cybersecurity executive Ori Eisen.
“There is a good chance that we will continue to receive funding,” Pascual said.