Cyber Security News

Hackers Attacking AI Agents To Hijacking Customer Sessions

Conversational AI platforms, powered by chatbots, are witnessing a surge in malicious attacks, which leverage NLP and ML are increasingly being used by businesses to enhance productivity and revenue.

While they offer personalized experiences and valuable data insights, they also pose significant privacy risks.

The collection and retention of user data, including sensitive information, raises concerns about data protection and the potential for breaches.

As the adoption of AI agents continues to grow, addressing these security challenges becomes paramount to ensuring the safe and effective use of conversational AI technologies.

Free Webinar on How to Protect Small Businesses Against Advanced Cyberthreats -> Free Registration

Conversational AI and Generative AI are two distinct branches of AI, each serving a specific purpose.

While Conversational AI focuses on two-way communication, understanding, and responding to human language, Generative AI specializes in creating new content based on learned patterns.

revealing personally identifiable information (PII)

Conversational AI is commonly used in chatbots and virtual assistants, while Generative AI finds applications in creative fields like text generation and image creation.

In essence, Conversational AI facilitates dialogue, while Generative AI innovates through content creation.

AI agents pose significant security risks, including data exposure, resource consumption, unauthorized activities, coding errors, supply chain risks, access management abuse, and malicious code propagation.

Conversational AI systems further exacerbate these risks by handling sensitive user data, which can be compromised if not properly secured.

To mitigate these threats, robust controls must be implemented to prevent data breaches, resource depletion, and unauthorized actions.

access specific customer sessions

In a recent breach, a threat actor gained access to a major AI-powered call center solution, compromising over 10 million conversations between consumers and AI agents, which exposed sensitive personally identifiable information (PII) that could be used for advanced phishing attacks and identity theft.

The compromised AI models may also have retained PII from their training data, posing additional risks, highlighting the need for robust security measures and continuous monitoring of AI systems to protect sensitive customer data.

Third-party AI systems pose a significant cybersecurity risk to enterprises due to potential data breaches and malicious data injection.

Attackers can exploit vulnerabilities such as unsecured credentials, phishing, and public-facing application exploits to gain unauthorized access to sensitive data and manipulate AI agent outputs.

targeting access tokens

The MITRE ATLAS Matrix provides a framework for identifying and addressing these risks. Enterprises must conduct thorough risk assessments before implementing third-party AI tools to mitigate potential negative consequences.

Resecurity highlights the criticality of a comprehensive AI TRiSM program to ensure the security, fairness, and reliability of conversational AI platforms.

Given the increasing reliance on these platforms, proactive measures like PIAs, zero-trust security, and secure communications are essential to mitigate privacy risks. 

Adversaries are targeting conversational AI due to their potential for data breaches and the vulnerability of the underlying technologies.

As these platforms evolve, it’s imperative to balance traditional cybersecurity with AI-specific measures to protect user privacy and prevent malicious exploitation.

Analyse Any Suspicious Links Using ANY.RUN’s New Safe Browsing Tool: Try It for Free

Aman Mishra

Aman Mishra is a Security and privacy Reporter covering various data breach, cyber crime, malware, & vulnerability.

Recent Posts

Zero-Trust Policy Bypass Enables Exploitation of Vulnerabilities and Manipulation of NHI Secrets

A new project has exposed a critical attack vector that exploits protocol vulnerabilities to disrupt…

2 days ago

Threat Actor Sells Burger King Backup System RCE Vulnerability for $4,000

A threat actor known as #LongNight has reportedly put up for sale remote code execution…

2 days ago

Chinese Nexus Hackers Exploit Ivanti Endpoint Manager Mobile Vulnerability

Ivanti disclosed two critical vulnerabilities, identified as CVE-2025-4427 and CVE-2025-4428, affecting Ivanti Endpoint Manager Mobile…

2 days ago

Hackers Target macOS Users with Fake Ledger Apps to Deploy Malware

Hackers are increasingly targeting macOS users with malicious clones of Ledger Live, the popular application…

2 days ago

EU Targets Stark Industries in Cyberattack Sanctions Crackdown

The European Union has escalated its response to Russia’s ongoing campaign of hybrid threats, announcing…

2 days ago

Venice.ai’s Unrestricted Access Sparks Concerns Over AI-Driven Cyber Threats

Venice.ai has rapidly emerged as a disruptive force in the AI landscape, positioning itself as…

2 days ago