The end of the scam: how to identify SMS fraud using artificial intelligence

SMS fraud, or phishing, has become one of the most common and disturbing threats to the public in Israel, as fake scam messages in the name of well-known companies and entities — such as Highway 6, Pango, the National Insurance Institute, and the Israel Electric Corporation — flood mobile devices every day. To deal with the growing phenomenon and prevent more victims from falling into the trap, artificial intelligence expert Dalia Baida presented in the 'HaOlam HaBoker' studio an accessible technological solution that allows fighting the scam with the help of AI.

The solution is based on setting up a personalized AI tool in the Gemini system, using a feature known as Gems.

"It's like our own chat with a very specific goal — in our case, it will be the scam detector, and we are also going to give it a prompt," Baida explains step by step.

In the setup process, a new Gem is opened — here with the name "Scam Detector" — and a prompt is entered that defines the AI as an expert agent for cyber and phishing detection.

For example, a suspicious SMS message regarding an alleged debt to Highway 6 will cause the Gem to display a very high risk rating for phishing. The problematic elements in the message can be the use of a shortened link, threats and the creation of immediate pressure, and an unofficial sender identity from a regular mobile phone.

So, what are the four simple steps to identify SMS fraud with the help of artificial intelligence?

1. Take a screenshot of the suspicious message.

2. Upload the image immediately to the personalized Gem in Gemini.

3. Perform an automatic analysis by the agent in real time.

4. Make an informed decision according to the results and arguments presented on the dashboard.

Implementing the tool offers a practical and daily solution not only for the users themselves, but mainly for elderly family members and parents, who may be exposed to falling victim to a scam.