In 2025, AI in everyday life was characterized by speed, scale, and constant change.

Organisations are now leveraging artificial intelligence and machine learning (AI/ML) across their entire operations to streamline workflows, automate decision-making, and boost productivity. AI is supporting development, communication, research, and operations at a pace that would have been unimaginable just a few years ago. However, this acceleration also brings with it an increasing number of trade-offs. In other words, more sensitive information is passing through more AI/ML applications, leading to reduced visibility and fewer safeguards.

As the AI footprint has expanded, organisations’ attack surfaces have grown, and threat actors have quickly followed suit over the past year. As barriers have lowered and attacks have become more realistic, they have also become faster and more reliable. At the same time, early signs of the misuse of agent-based and semi-autonomous AI indicate a shift in how threats are evolving. Organisations are now grappling with a growing range of risks, from shadow AI and embedded AI to hallucinations and unprotected private models.