Cybercriminals never sleep. They craft messages that mimic urgent bank alerts or irresistible deals, slipping past even wary eyes. Now, Malwarebytes fights back from inside Anthropic’s Claude. The company’s free connector, launched April 29, 2026, lets users paste suspicious links, phone numbers, emails, or screenshots directly into chats for instant verdicts. Malicious. Suspicious. Safe. Unknown. Each comes with clear next steps, powered by Malwarebytes’ threat databases that shield millions of devices.
Marcin Kleczynski, Malwarebytes’ founder and CEO, put it bluntly: “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.” His firm, long known for malware hunters, extends that vigilance to conversational AI. Last year, research showed 66% of people struggle to spot scams. Over 19% of Malwarebytes Scam Guard users dodged high-risk traps—ones risking $1,000 or more, or worse. This tool meets fraud where it thrives: texts, social posts, calls.
Setup takes seconds. Open Claude. Hit Customize > Connectors. Tap the plus, browse for Malwarebytes, connect. No account needed. Paste a URL from a shady text: “Malwarebytes, is http://txdmv.gov-mbjn.vip/track safe?” It scans against phishing lists, flags impersonations of real sites like state DMVs. Phone numbers get checked for scam ties, revealing carrier, region, known fraud patterns. Emails? Domain legitimacy via WHOIS—new registrations scream caution. And it handles batches. Drop a full message with links, numbers, addresses. Claude extracts, analyzes all at once. Even screenshots or docs trigger auto-checks, as detailed in Malwarebytes’ help guide.
Verdicts drive action. Malicious means block it—confirmed threats. Suspicious? Proceed carefully; maybe dig into WHOIS for a fresh domain from a dodgy registrar. Unknown prompts extra scrutiny. Users report flags straight from chat. Only the suspect item goes to Malwarebytes’ intelligence team—no full convos, no personal data. That feedback loop sharpens defenses for everyone, as Malwarebytes explained in its blog.
This builds on February’s ChatGPT plug-in. Claude’s edge? Superior reasoning pairs with threat intel for context-rich calls, spotting AI-forged lures that blend hype and fear. Digital Trends called it a potential lifesaver against costly errors. BetaNews highlighted batch power: “Users can paste full suspicious messages… for collective checks.” Privacy holds; queries send isolated data points.
But AI draws its own wolves. Just weeks ago, fake Claude sites peddled PlugX malware via trojanized installers, mimicking legit paths with typos like “Cluade.” Malwarebytes’ own researchers dissected the chain: ZIPs dropping signed G Data updaters for DLL sideloading, granting remote access. Cybernews echoed the threat, tying it to Claude’s 290 million monthly visits. The connector counters this—users query odd downloads or sites mid-chat.
Industry pros see broader shifts. Connectors turn Claude into a security hub, but risks lurk. They grant external access to chats; vet them. Still, for pros juggling alerts amid workflows, this embeds checks without tab-switching. Imagine dissecting a client phishing wave: paste samples, get verdicts, report en masse. Kleczynski again: “We’re putting enterprise-grade… detection into the hands of anyone… right when—and where—they need it most.” PR Newswire’s release spells out the play.
Scams cost billions yearly. Americans alone lose heaps to social traps. This isn’t foolproof—no tool is. Unknowns persist; human judgment rules. Yet it arms the front line. Pros gain speed. Casual users dodge hooks. And as AI chats deepen daily reliance, such guards become table stakes. Malwarebytes in Claude. Your new reflex against the grind.
WebProNews is an iEntry Publication