MAD-CAT “Meow” Tool Sparks Real-World Data Corruption Attacks

MAD-CAT, a Meow Attack Data Corruption Automation Tool, is described as a custom-built adversarial simulation that shows how attackers can corrupt data across multiple database platforms at once. Researchers note that MAD-CAT can operate across six real-world targets—MongoDB, Elasticsearch, Cassandra, Redis, CouchDB, and Hadoop HDFS—using bulk CSV campaigns to strike entire database ecosystems in coordinated fashion.

The tool imitates the 2020 Meow attack by replacing string and numeric fields with ten-character random alphanumeric strings followed by “-MEOW”, mirroring the signature from that period. Unlike earlier single-target exploits, MAD-CAT represents a significant escalation by enabling simultaneous data destruction across several platforms, potentially crippling organisations within minutes.

The article highlights that even years after the initial campaigns, compromised databases persist, underscoring ongoing risks from misconfiguration and weak authentication, while advocating for default authentication, credential rotation, network segmentation, and robust backups as essential defensive measures.