Cybersecurity

"ChatGPT, is this a scam?": now Norton helps with the answer

Norton's Genie anti-scam assistant responds directly in ChatGPT

by Marco Trabucchi

17 March 2026

3' min read

Translated by AI

Versione italiana

3' min read

Translated by AI

Versione italiana

Artificial intelligence has become our favourite oracle, we ask it for everything: recipes, translations, summaries of contracts, advice on how to answer a difficult email. But there is one question that, in recent months, has become increasingly common in the chats of AI assistants around the world: "Is this message legitimate?". A text message from the bank asking you to click on a link. An email warning of an account suspension. A WhatsApp message that seems to come from a colleague, or even the boss.

The phenomenon is real and growing: people, when faced with doubt, are turning to AI assistants as they would once have called a computer-savvy friend. It is precisely on this change in user behaviour that Norton - the historic consumer cybersecurity brand, part of the Gen Digital group (listed on the Nasdaq) - has built its most recent innovation: its anti-fraud assistant Genie integrated directly within ChatGPT.

The rationale is one of immediacy: instead of convincing users to open a new security app, Norton brings protection to the most used AI environment. After activating the Norton app in ChatGPT's app directory, users can mention @Norton in the chat and paste the suspicious message, an email, link or even an image, receiving an immediate risk analysis in response. The integration is available to ChatGPT users including Free plans, with a potential audience of millions globally.

"People are already asking ChatGPT whether they should click, pay or respond," said Leena Elias, Chief Product Officer at Gen Digital. "With Genie integrated into ChatGPT, we are extending Norton's scam analysis and advice directly into those conversations."

Unlike traditional link-checking tools - which check whether a URL is already known to be malicious - Genie analyses the overall context of the message: the language used, the psychological tactics employed (urgency, fear, authority), the characteristics of the originating domain. An approach that reflects the very nature of modern scams: increasingly sophisticated, contextualised, and difficult to unmask without critical reading.

90% of threats come from scams and phishing

According to the Gen Threat Report, in 2025 more than 90 per cent of threats aimed at consumer users come from scams, phishing and malvertising. Not from viruses in the classical sense, but from deception: messages designed to seem normal, convincing, urgent.

In this scenario, the traditional response, such as an antivirus on the computer, a spam filter in the inbox, remains necessary but no longer sufficient. Phishing attacks - the computer scam based on social engineering - are arriving everywhere: on WhatsApp, via SMS, in Instagram DMs, in business emails, even in comments on social networks and even from deepfakes. Protecting oneself means developing critical reading skills that not everyone possesses and that the factor of fatigue or distraction can easily undermine.

AI as second opinion: a paradigm shift

There is also a cultural element to this evolution. When a person asks an AI assistant "is this a scam?", they are replicating a very familiar behaviour: that of asking for a second opinion. In the past, this was done with the more tech-savvy colleague, friend or 'geeky' relative. Today, that role, for millions of users, is increasingly being played by digital assistants.

AI thus becomes a quick decision filter: a tool to turn to before clicking on a link, making a payment or replying to a suspicious message. It does not replace human judgement, but goes alongside it, offering an immediate assessment when doubt emerges in the middle of a conversation, on a phone or in front of an inbox.

In this scenario, Norton has chosen to insert itself into the decision-making process, bringing its analysis tools into the conversation itself. "People can now assess their cyber risks quickly and conveniently, at any time," explained Silvia Signorelli, Senior Marketing Manager Italy and Balkans at Gen Digital. The result is a form of security assistance that does not require technical expertise or the installation of new software: just type in a prompt.