Google Implements AI To Fight Scams

On May 8, Google announced it had ramped up its use of AI tools to protect Chrome and Android users from potential scams. These tools now include the Gemini large language model (LLM) to detect potentially dangerous websites and even warn Android users of suspicious texts and calls.

Google started implementing AI as part of its privacy protection tools as far back as 2020, when it debuted its Enhanced Protection mode of Safe Browsing for Chrome.

However, this announcement marks the first time Google has implemented an LLM as part of its security catalogue. On desktop devices, Chrome now runs an in-device variant of Google’s standard Gemini model called “Gemini Nano.” Nano will scan a website’s content and return its findings to Google Safe Browsing, which will tell the user if a site is likely to be part of a phishing scam or similar.

“The on-device approach provides instant insight on risky websites and allows us to offer protection, even against scams that haven’t been seen before,” reads the company’s announcement blog post. “Gemini Nano’s LLM is perfect for this use because of its ability to distill the varied, complex nature of websites, helping us adapt to new scam tactics more quickly.”

The company has already started using this AI approach when it comes to remote tech support claims.

Phishing schemes are especially dangerous since they use social engineering as a first vector for attack, instead of more technical means, which turns each user into the highest point of vulnerability. Many prominent and recent cyberattacks started by using phishing tactics, including a recent hack that affected over 32 Chrome extensions.

The company will also extend its AI-powered safety features to Android devices in the form of the “Scam Detection in Google Messages and Phone by Google” tool. This tool will scan messages and even call content to detect suspicious patterns and warn users of a probable scam.

According to Google, since these security models will run on-device, the personal information of users will remain private and secure.

Google has put a lot of stock in its Gemini models lately, hoping to expand its functionality and popularity. Earlier this year, the tech giant introduced a significant memory upgrade to the model.