Generative AI is rapidly influencing a wide range of fields — cybersecurity included. As large language models (LLMs) become more accessible and capable, they open new possibilities for enhancing security operations while also introducing novel risks.
Within cybersecurity, AI is already being adopted to support use cases such as incident triage, threat analysis, secure coding assistance, and knowledge management. However, the level of maturity varies significantly across organisations, with many still in the exploratory or pilot phase. Adoption is often driven by the promise of increased efficiency, but concerns around governance, data sensitivity, and trustworthiness remain key barriers.
This duality defines the current landscape. While generative AI empowers defenders to automate workflows and analyse threats at scale, adversaries increasingly weaponize these tools to craft convincing phishing campaigns, generate malicious code, and exploit vulnerabilities faster than ever. For cybersecurity teams, the challenge lies in balancing innovation with vigilance—harnessing AI’s potential without underestimating its risks.
CERT-EU's Perspective
CERT-EU presents practical guidance to help cybersecurity organisations navigate the adoption of generative AI responsibly. Drawing from both internal experience and ongoing monitoring of the cybersecurity landscape, we aim to share practical insights on how these technologies can be integrated effectively and securely.
Whether you are exploring the use of LLMs internally or assessing the impact of these technologies on your threat landscape, this guidance offers a cybersecurity-focused perspective to support informed and responsible decision-making.