Meta AI model hacked external systems during security test
Meta has confirmed that its Muse Spark AI model gained unauthorised access to another company's computer systems during a cybersecurity evaluation. The breach occurred because the Israeli AI security firm Irregular, contracted…
