Analysis ·
Cybersecurity and agentic AI: best practices to adopt
Original title, in French: “Cybersécurité et IA agentique : les bonnes pratiques à adopter”
- Publication
- Journal du Net (France)
- Author
- Adrien Gendre
- Published on
- 30 September 2026 · read in the page structured data
Summary
The article explains how agentic AI, by being given broader access to company data and systems, creates new cybersecurity risks. Attackers exploit these agents through prompt injection to extract data, and are developing ransomware into a so-called "3.0" version that corrupts or falsifies data rather than simply encrypting it. According to a study cited, 61% of CISOs believe AI has directly increased the risk of ransomware attacks. The article recommends applying the principle of least privilege, continuously monitoring the behaviour of AI agents, and adopting a Zero Trust architecture with strong authentication (FIDO2, Passkeys) to secure AI adoption in business.
Summary based on the full text of the article.
Read the article on Journal du Net →Figures cited
-
61 %
Share of CISOs who believe AI has increased the risk of ransomware attacks
CISOs, study not precisely dated in the article
« 61 % des RSSI ont déclaré que l'IA avait directement augmenté le risque d'attaques par ransomware (1). »
Classification
- Country
- Worldwide scope
- Company size
- All sizes
- Industries
- Software vendors, IT services companies, telecoms · All industries, or industry not specified
- Keywords
- agentic AI, cybersecurity, ransomware, zero trust, access governance, identity management
Entry written by Marsanne Studio from the cited article, with the help of an AI model (claude-sonnet-5). The rights to the article and its illustrations belong to Journal du Net and its author. Entry translated from French by Claude Sonnet 5.5.