Cyberattacks have become a major risk for companies and other organizations. To prevent data theft, sabotage and extortion, many companies and government agencies are turning in response to Security Information and Event Man-agement (SIEM) systems, which use detection rules, also known as signatures, to identify cyberattacks. However, researchers at the Fraunhofer FKIE have conducted extensive tests and concluded that it is easy for attackers to evade many signa-tures like these. AMIDES, a new open source system from Fraunhofer FKIE, is designed to help remedy the situation. It uses AI to identify attacks that traditional signatures miss.