---
title: 8 security practices in AI that companies should avoid, according to CrowdStrike
url: https://www.elseif.net/8-security-practices-in-ai-that-companies-should-avoid-according-to-crowdstrike
published: 2026-10-22T15:03:43+00:00
language: en
section: Agents
source: https://itforum.com.br/noticias/crowdstrike-8-praticas-seguranca-em-ia/
organizations: CrowdStrike
publisher: elseif
---

# 8 security practices in AI that companies should avoid, according to CrowdStrike

Experts warn that artificial intelligence is a new attack surface that requires continuous attention and employee education. The use of artificial intelligence (AI) by companies and teams brings a new layer of risk to cybersecurity. CrowdStrike published a list on Monday of eight practices that organizations of all sizes should avoid when operating tools and AI agents. This publication is part of Cybersecurity Awareness Month, celebrated annually in October. According to the company, AI has become an attack surface that requires ongoing attention and basic security hygiene.

The Vice President of Sales Engineering for Latin America at CrowdStrike states that "AI can be a powerful ally, but it also expands the attack surface. Many risks begin with everyday decisions made without due care. Organizations need to remain vigilant and continually educate their employees about these dangers. Those that will succeed are not those that slow down AI, but those that advance rapidly without losing control, and visibility and protection become fundamental items in this journey."

CrowdStrike recommends that organizations avoid certain behaviors when working with AI tools and agents. CrowdStrike is listed on Nasdaq under the ticker CRWD and focuses on endpoint protection, cloud workloads, identity, and data through its Falcon platform.