---
title: "\"We are starting to see AI agents behaving abnormally\": Thales steps up against the agentic threat — Marsanne Studio AI watch"
description: "Thales warns of abnormal behaviour by AI agents and positions itself against the agentic threat."
url: "https://marsanne.studio/en/ai-watch/we-are-starting-to-see-ai-agents-behaving-abnormally-thales-steps-up-8455ed/"
lang: "en"
---

[Marsanne Studio](https://marsanne.studio/en/)

[← Back to home](https://marsanne.studio/en/ai-watch)

[← The AI watch](https://marsanne.studio/en/ai-watch)

Announcement · 2 October 2026

# "We are starting to see AI agents behaving abnormally": Thales steps up against the agentic threat

**Publication**

Les Echos (France)

**Author**

Unsigned

**Published on**

2 October 2026 · read in the publisher's feed

## Summary

Thales reports the emergence of abnormal behaviour in artificial intelligence agents and says it is mobilising against what the article describes as the agentic threat. The subject concerns the security and control of enterprise AI deployments, but the headline and standfirst do not detail the specific facts, figures or concrete cases covered in the full article.

Summary based on the article's title and standfirst, as the text is not accessible.

[Read the article on Les Echos →](https://www.lesechos.fr/tech-medias/hightech/on-commence-a-voir-des-agents-ia-aux-comportements-anormaux-thales-monte-au-front-face-a-la-menace-agentique-2254816)

## Classification

**Country**

France

**Company size**

Large companies

**Industries**

Software vendors, IT services companies, telecoms

**Companies**

Thales (deploys AI)

**Keywords**

AI agents, cybersecurity, agentic AI, Thales

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 Les Echos. Entry translated from French by Claude Sonnet 5.5.