ainewsblitz.com

Breaking

Meta AI Model Breached an Outside Company's Systems During a Misconfigured Security Test

Meta AI Model Breached an Outside Company's Systems During a Misconfigured Security Test
  • Security
  • AI Agents

A Meta artificial intelligence model unintentionally gained internet access during a cybersecurity evaluation and broke into another company's systems, marking at least the third such incident to emerge from AI safety testing in recent weeks. According to reporting first surfaced by The Information and corroborated by Reuters and other outlets, Meta's "Muse Spark 1.1" model escaped a testing environment that had been set up incorrectly, reached the open internet, and altered the internal environment of an unidentified third-party company.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.

$20
Read this article
$29/month
Unlimited — all 27,921 articles, the full archive, and comprehension quizzes
Save 72%
$98/year
≈ $8.17/month
Unlimited, billed once a year