Cybersecurity entrepreneur Josh Lospinoso discusses the threats and potential of artificial intelligence (AI) in military operations, highlighting the importance of securing existing systems and approaching AI development responsibly.
Josh Lospinoso, a cybersecurity entrepreneur and former Army captain, discussed the threats and potential of artificial intelligence (AI) in military operations. He highlighted two main threats to AI-enabled technologies: theft and data poisoning, which refers to adversaries crafting misleading data to manipulate AI systems. Although data poisoning is not widespread, notable incidents like Microsoft’s Twitter chatbot, Tay, being influenced by malicious users demonstrate its potential impact. Lospinoso emphasized the use of AI in cybersecurity, such as email filters and malware detection, but also highlighted the vulnerabilities in existing military software systems. He called for securing these systems and addressing technical debt before adding new AI capabilities. Lospinoso expressed concerns about rushing AI products to market without adequate security measures and urged for a responsible and secure approach to AI development. He emphasized that AI algorithms are not yet ready for making decisions in lethal weapon systems.
