Norton Parent Company Proposes New Concept: AI Agents Need VPNs Too

When AI Agents Go Online, Who Protects Their 'Privacy'?

You're probably used to using a VPN to protect your online privacy, but have you ever considered that your AI agent might need one too? Gen Digital, the parent company of Norton and Avast, is seriously exploring this question — and the answer may be more important than you think.

As AI agents transition from lab concepts to everyday applications, they are taking over an increasing number of online tasks on behalf of humans: browsing the web, booking flights, managing emails, and even conducting financial transactions. This means the digital footprint left by AI agents on the internet is growing rapidly, while security protections surrounding them remain virtually nonexistent.

Security Threats Facing AI Agents

Traditional cybersecurity systems are designed around human users. Firewalls, antivirus software, VPNs, and other tools have always been built to protect the person sitting in front of the screen. However, AI agents operate in a fundamentally different way: they can autonomously access multiple websites and APIs, transmit sensitive data, and authenticate on behalf of users — all with little to no human intervention.

This autonomy introduces entirely new security risks:

• Data Breach Risk: AI agents may transmit users' personal information, payment credentials, and private data while performing tasks. If communication channels are unencrypted, this information is highly vulnerable to man-in-the-middle attacks.
• Identity Spoofing Attacks: Malicious actors could impersonate legitimate service endpoints to trick AI agents into submitting sensitive information.
• Behavioral Tracking and Profiling: The network request patterns of AI agents can be tracked and analyzed to infer users' behavioral habits, business intentions, and even financial status.
• Prompt Injection Attacks: While browsing the web, AI agents may encounter malicious instructions embedded in webpage content, manipulating them into performing unintended actions.

Gen Digital's Strategic Vision

Gen Digital, one of the world's largest consumer cybersecurity companies, owns well-known brands including Norton, Avast, LifeLock, and AVG. The company has keenly recognized that the rise of AI agents will fundamentally transform the cybersecurity protection paradigm.

Providing VPN services for AI agents essentially means establishing an encrypted "secure tunnel" for them. Through a VPN, an AI agent's network traffic is encrypted and its IP address is concealed, significantly reducing the risk of data interception and tracking. This protects not only the AI agent itself but also the human user it represents.

From a business perspective, this concept also signals Gen Digital's strategic shift from traditional "protecting human users" to "protecting AI users." As major tech giants roll out AI agent products — from OpenAI's Operator to Google's Project Mariner to Anthropic's Claude computer use capabilities — the number of AI agents could soon surpass the number of human internet users, representing a massive emerging market.

Industry Chain Reaction

Gen Digital's move could trigger a chain reaction across the entire cybersecurity industry. As AI agents become one of the internet's primary "inhabitants," existing security infrastructure will need to be re-examined:

Authentication systems will need to expand to accommodate identity verification for non-human entities; traffic monitoring systems will need to distinguish between human traffic and AI agent traffic; and privacy protection regulations may need to be updated to clarify compliance requirements for data collection and transmission by AI agents.

Notably, the security needs of AI agents extend far beyond VPNs alone. In the future, we may see an entirely new category of security products designed specifically for AI agents, including firewalls, intrusion detection systems, and behavioral auditing tools.

Outlook: The Co-Evolution of Security and Intelligence

We are standing at a critical turning point. AI agents are evolving from "tools" into "digital proxies" that will act, transact, and make decisions on the internet on behalf of users. In this process, security protection cannot be an afterthought — it must be embedded into the design and deployment of AI agents from the very beginning.

Gen Digital's pioneering concept of "equipping AI agents with VPNs" is still in the conceptual stage, but it sends a clear signal: the core battlefield of next-generation cybersecurity will no longer be solely about protecting humans, but about protecting the entire digital ecosystem composed of both humans and AI agents.

For everyday users, this is also a topic worth paying attention to early. As you delegate more and more tasks to AI agents, remember — they represent you online, and their security is your security.