SUMMER 2026

28 THE DISTRIBUTOR’S

28 THE DISTRIBUTOR’S LINK Joe Dysart Joe Dysart is an Internet speaker and business consultant based in Thousand Oaks, California. A journalist for over 20 years, his articles have appeared in more than 40 publications, including The New York Times and The Financial Times of London. During the past decade, his work has focused exclusively on ecommerce. Telephone: 631-759-7207 or email: joe@breakingintech.com WHY EVERY FASTENER DISTRIBUTOR NEEDS TO DOUBLE-DOWN ON SOFTWARE SECURITY NOW Fastener distributors across the globe will soon be facing a very special nightmare created by a new AI model, which has the proven ability to discover thousands of new security vulnerabilities in hundreds of the most commonly used software programs on the market. Dubbed Mythos and slated for release later this year by its maker Anthropic, the AI model was deemed so dangerous when first developed, Anthropic put the brakes on its release and consulted with many of the world’s top companies to find a way to literally protect the world from the AI. The result: Project Glasswing, a joint effort by the likes of Apple, Google, Microsoft, Amazon and similar to run Mythos against the world’s most popular software programs - and plug the torrent of security holes Mythos found before hackers use the same software to find those same vulnerabilities. So far, the depth of security holes that Mythos has found is beyond disturbing. Mozilla, maker of the wildly popular Firefox browser, for example, discovered 271 previously unknown security vulnerabilities in its software with Mythos - and had to race to fix them all before Mythos is released to the world. Cloudflare, a major cybersecurity provider, has uncovered more than 2,000 security holes with Mythos, including 400 that have a severity level of high or critical. And even worse: A security test of 1,000+ Open Source software apps by Mythos - or apps that are free to use by businesses across the world - found 23,019 security gaps in those apps, including 6,202 security problems with a severity level of high or critical. Says Jason Turner, chairman, seQure, an AI-powered security provider, warning of things to come: “Mythos-type CONTRIBUTOR ARTICLE A SOON-TO-BE-RELEASED AI MODEL IS EXPECTED TO WREAK HAVOC ON SOFTWARE SECURITY SYSTEMS WORLDWIDE attacks represent new superpowers for attackers.” Short-term, Project Glasswing has been a somewhat helpful fix and has helped the world’s major software makers to get somewhat of a headstart on hackers - even though they have a long way to go. But long-term, the impact of the Mythos release is anyone’s guess. Granted, major software programs like Firefox, Microsoft Office, Google Workspace and similar have been (hopefully) safeguarded by Glasswing. But unfortunately, those efforts won’t deter hackers from unleashing the full power of Mythos -- and similar AI emerging from other AI players -- against less popular software programs, older software programs and custom software programs that reside on the computer networks of fastener distributors and other businesses across the globe. CONTINUED ON PAGE 102