Meta AI's Security Sandbox Blunder Exposes External Firm to Breach Meta AI, a subsidiary of Facebook, conducts security testing as part of its artificial intelligence development process.
This involves simulating attacks on systems to identify vulnerabilities and strengthen defenses.
In a recent incident, Meta AI made a critical error during security testing, breaching an external firm's systems through a sandbox testing environment.