Mr. Goldstein is a professor at Vanderbilt University who specializes in cybersecurity and artificial intelligenceAnthropic recently sent a shock wave through the cybersecurity world when it said its new artificial intelligence model, Claude Mythos, had exhibited an extraordinary ability to find previously unknown vulnerabilities in software — a hacker’s fantasy. Concern over the tool’s power caused Anthropic to restrict its release mainly to bigger companies, allowing them time to secure their software. What is everyone else supposed to do? Smaller companies, organizations, nonprofits and regular people are just as much at risk as larger companies. But they most likely lack...