Vulnerabilities, AI Compete for Software Developers' Attention

The report analyzes how rapid adoption of AI coding assistants is boosting developer productivity but outpacing security, contributing to longer vulnerability remediation times in open source projects and elevating software supply chain risk. It highlights research demonstrating threats like LLM poisoning and hallucination-driven steering to malicious packages, notes developers’ limited trust in AI-generated code security, and argues that organizations must implement guardrails and automated security tools while AI models mature to produce more secure code.