Enterprises neglect AI security – and attackers have noticed

IBM’s 2025 Cost of a Data Breach Report warns that rapid enterprise AI adoption is outpacing security and governance, with 13% of surveyed organizations reporting AI-related security incidents and 97% of those lacking proper AI access controls; supply chain and SaaS providers were the leading sources of intrusions, while shadow AI, missing governance (87%), infrequent audits, and limited adversarial testing exacerbate exposure—signaling attackers are beginning to exploit this oversight gap and that AI security must be treated as foundational.