AI in Cloud: Critical Cybersecurity Flaws Leave Organizations Vulnerable

Organizations leveraging artificial intelligence in cloud environments are frequently risking their cybersecurity by making significant configuration errors. A Tenable report reveals that 91% of those surveyed had deployed Amazon SageMaker with root access enabled by default, posing serious risks if compromised, as root access can allow manipulation of critical systems. The report highlights the trend of excessive permissions and public exposure as significant vulnerabilities, alongside the "Jenga concept" where each service inherits risks from others, leading to potential cascading failures. To mitigate these risks, organizations need to proactively reassess their security settings rather than relying on default configurations.

Opens on Dark Reading