If you’ve ever pasted a sketchy text into an AI chatbot hoping for a straight answer on whether it’s a scam, you already understand the problem: generalist AI is pretty good at reasoning through suspicious-looking language, but it doesn’t have the threat intelligence underneath to back it up. Norton is fixing that.
As of June 30, the Norton Genie scam detector is available as a native connector inside Claude, across all subscription tiers. The integration lets you paste in a dodgy message, link, email, or screenshot and get an analysis that draws on Norton’s actual security infrastructure including URL reputation databases, redirect tracing, domain history, and phishing pattern recognition, rather than just a language model’s best guess. Norton first rolled out the same connector for ChatGPT back in February, so if you’ve already been using it there, Claude support is the natural next step.
For Claude, the setup is straightforward: log into Claude, find Norton in the Connectors (MCP) section of the settings sidebar, enable it, and you’re done. For ChatGPT, you go through the App Directory and tag @Norton in any conversation once it’s connected. Either way, you just describe what you’re looking at and ask.
What it does that neither chatbot can pull off alone
The honest pitch here isn’t that these AI assistants were useless for scam questions before. It’s that there’s a specific category of threat, the kind that looks totally legitimate on the surface, where a language model’s pattern-matching runs out of runway fast. A phishing site that mimics your bank exactly, a shortened URL that redirects three times before landing somewhere malicious, a fake e-shop with plausible reviews: Claude or ChatGPT can flag the vibes, but Norton can follow the links (if allowed) and check the receipts.
That matters more than it used to. Gen’s own threat research found that more than 90 percent of threats targeting consumers in 2025 came from scams, phishing, and fake ads, and their researchers are tracking more than 1,000 new fake e-shops appearing every day on average. The FTC reported that Americans lost billions to social media scams alone last year, up eightfold from 2020. Tax scam victims, per Norton’s own research, lost an average of $8,401 each: These aren’t abstract numbers, they’re the threat model that makes a tool like this worth having.
One thing worth knowing: Norton doesn’t just scan links but it also reads the broader context of a message, the urgency language, the impersonation tactics, the requests for sensitive information and layers that on top of the URL and domain checks. That’s the gap a standalone chatbot can’t close on its own and it’s a reasonable reason to have both running together.
On pricing, Norton’s standalone subscription plans in the US currently start at $2.50 a month for AntiVirus Plus (one device), going up to $4.17 a month for Norton 360 Deluxe which covers five devices and adds a VPN, dark web monitoring, and parental controls. The connector itself doesn’t cost extra on top of whichever Claude or ChatGPT plan you’re already on, which makes it a low-friction add for anyone who already subscribes to Norton.
See Norton’s Plans and Pricing
It’s worth noting the family connection: Norton is part of Gen Digital, the same parent company behind Avast, which recently overhauled its own consumer security app around a modular, build-what-you-need model. Gen is clearly pushing its various brands into AI-native contexts wherever it can and this connector is the clearest example yet of what that strategy looks like in practice.