Malwarebytes is adding its scam detection tools directly into AI chats, starting with a new connector for Claude that lets users check links, phone numbers, and messages without leaving the conversation.
Online scams increasingly rely on language tricks, fake urgency, and convincing messages rather than traditional malware files. Many attacks now arrive through texts, emails, social media messages, or phone calls, which makes quick verification tools more useful than traditional antivirus alone.
The new integration, called Malwarebytes in Claude, connects the company’s threat intelligence system to Anthropic’s Claude assistant. It runs checks inside the chat instead of sending users to a separate website or security tool, and allows users to paste links, phone numbers, email addresses, or domains directly into the AI to check whether they are safe. The system then analyzes those items using its threat intelligence databases and returns a verdict.
Each check produces one of four results: Malicious, Suspicious, Safe, or Unknown. The response also includes context explaining why something was flagged, which will help users decide whether to ignore, block, or report the content.
The connector also supports checking multiple items at once, which can come in handy when a message includes several links or contact details. Users can paste a full suspicious message and have Claude analyze the contents together instead of running checks one at a time.
Domain verification allows users to look up registration details through WHOIS records. This can reveal when a domain was created and whether it matches the expected organization, which can expose recently created scam websites.
Users can also report suspicious links or numbers directly through the connector. Those reports go to Malwarebytes’ threat intelligence team, where they can be analyzed and added to detection databases used across its security tools.
The company says data from its Scam Guard tool shows more than 19 percent of users avoided high-risk scams, including incidents that could have resulted in losses of $1,000 or more.
“The use of AI has opened up new areas for cybercriminals to target, and combined with manipulative language or images, no one is too smart to fall victim,” Marcin Kleczynski, founder and CEO of Malwarebytes, said.
“We are tackling scams from every angle, helping people answer the question, ‘is this a scam’ and providing clear, easy-to-understand support on what to do next,” he added. “By integrating Malwarebytes with Claude, we’re putting enterprise-grade threat and scam detection into the hands of anyone who needs it, right when — and where — they need it most.”
Setting up the connector involves opening Claude, navigating to the connectors section, and selecting Malwarebytes from the available list. No account is required, so you can get started straightaway.
Privacy handling is built around sending only the items that need analysis, such as links or phone numbers. Full conversations and personal details aren’t stored.
What do you think about Malwarebytes bringing real-time scam detection into Claude? Let us know in the comments.
