Cybersecurity in the Age of AI-Generated Instant Software

AI is transforming software development by enabling "instant software"—custom applications generated on demand and deleted after use, alongside traditional long-term software. Attackers are already leveraging AI to automate vulnerability discovery and exploitation, with both government and criminal hackers using AI tools to target systems, including open-source libraries and IoT devices like internet-connected cars and industrial control systems. Defenders are adopting AI to find and patch vulnerabilities, though current AI-generated code remains insecure, as demonstrated by projects like OpenClaw. The effectiveness of AI in patching legacy software—particularly in IoT and industrial systems—remains uncertain due to testing and deployment challenges. Future scenarios include self-healing networks where AI agents continuously scan and patch vulnerabilities, though policy changes would be required to enable automated updates. Attackers may shift focus to social engineering, deepfakes, or exploiting AI system vulnerabilities like prompt injection.