AI speeds up discovery of software flaws and points to a wave of cyberattacks in 2027
Artificial intelligence is detecting vulnerabilities in the software that runs banks, hospitals, power grids and retailers at speeds far above human capacity. In April, Anthropic presented Claude Mythos, a model that found a 27-year-old flaw in OpenBSD, and in its first month the Glasswing program found more than 10,000 vulnerabilities in essential software. A record 48,185 vulnerabilities were published in 2025 and more than 100,000 have already been surpassed in 2026. The average time between disclosure of a flaw and its exploitation fell from 120 to 80 days.
In 45 seconds
- Anthropic's Claude Mythos found a 27-year-old flaw in OpenBSD; the Glasswing program detected more than 10,000 vulnerabilities in essential software in its first month.
- Published vulnerabilities rose from a record 48,185 in 2025 to more than 100,000 in 2026, an explosion experts attribute to AI.
- The average time between disclosure of a flaw and its exploitation dropped from 120 to 80 days.
IASummary generated by artificial intelligence from a story by El Tiempo. It may contain errors: the original source prevails.




Comments
Be respectful. Comments reported by several people are hidden until a moderator reviews them.
You do not need an account to read, only to comment. Sign in with a code sent to your email or with your Google account. We only show the name you choose, never your email.