MEDIUMAi
Global

'Turf War' Between Claude Agents Leads to Self-Replicating Malware

·Source: Dark Reading

Updated:

Executive Summary

Three testing models with the same goal but different directives engaged in "increasingly aggressive" territorial attacks on one another, according to Anthropic.

Analysis

Three testing models with the same goal but different directives engaged in "increasingly aggressive" territorial attacks on one another, according to Anthropic.
Source Attribution

Originally published by Dark Reading on Aug 17, 2026.

Related Threats

CRITICALAi

Microsoft finally patches critical one-click Copilot vulnerability, almost eight months after learning of it

Almost eight months after confirming a critical security vulnerability within the personal version of its AI assistant, Copilot, Microsoft on Tuesday issued a patch to close the hole, which relies on an LLM’s inability to distinguish the data in a query from an instruction. The CoSnitch hole was discovered by Varonis, and marked the third Copilot bug that Varonis has reported to Microsoft this yea

CSO Online
MEDIUMAi

Westpac to reshape software engineering with Amp Frontier Corporation

Westpac has contracted with San Franciscop-based Amp Frontier Corporation to build a suite of AI agents for software engineering tasks.

Finextra
MEDIUMAi

Perplexity Builds Guardrails to Rein in Rogue AI Agents

<img src="https://ismg-cdn.nyc3.cdn.digitaloceanspaces.com/articles/perplexity-builds-guardrails-to-rein-in-rogue-ai-agents-image_small-9-a-32596.jpg" align=right hspace=4><b>Open-Source Numbat Blocks Agent Actions That Violate Enterprise Security Policies</b><br>Perplexity designed its open-source tool Numbat tool to detect and block AI agents from stepping outside enterprise safety policies. The

Bank Info Security