AWS intruder pulled off AI-assisted cloud break-in in 8 mins

A cybercriminal used AI assistance to break into an AWS cloud environment and gain administrative privileges in under 10 minutes, compromising 19 AWS identities after stealing credentials from public S3 buckets. The attacker leveraged large language models to automate reconnaissance, privilege escalation, and code writing, then performed "LLMjacking" by accessing cloud-hosted AI models for unauthorized compute usage. Security researchers identified AI involvement through Serbian code comments, hallucinated AWS account IDs, and references to non-existent GitHub repositories, highlighting how criminals increasingly use AI to accelerate cyberattacks.