The global cybersecurity gap deepens as AI-powered attacks surge
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Anthropic unveiled Mythos Preview, a model that flagged thousands of vulnerabilities across every major OS and web browser, but limited access to about 40 firms, excluding most central banks and governments.
- CrowdStrike reported AI‑enabled cyberattacks surged 89% in 2025, and Zero Day Clock notes the median time from vulnerability disclosure to exploitation fell from 771 days in 2018 to just four hours by 2024.
- OpenAI's tools were used by a North Korean hacker group alongside Cursor to steal $12 M in cryptocurrency, showing how attackers weaponize AI faster than defenders can patch.
Why it matters: Governments and midsize firms lose protection while AI firms like Anthropic gain monopoly; with AI attacks up 89% and $12 M crypto thefts, the rushed patch cycle forces firms to spend billions on AI‑driven defenses, widening the digital divide.
