How to Protect Your Privacy When Using AI Tools: Advice from Proton’s CEO

AI tools like ChatGPT, Google Bard, and Microsoft Copilot are now part of daily life for millions of people. They help us draft emails, summarize documents, brainstorm ideas, and even code. But every time you paste a block of text or ask a question, you are handing over data that can be stored, analyzed, and possibly used to train future models.

Proton, the company behind ProtonMail and ProtonVPN, has built its reputation on strong encryption and user privacy. In a recent interview with Spiceworks, Proton’s CEO shared his view that privacy in the AI era is achievable—but he also pointed to one concern that keeps him up at night.

What happened

The CEO, whose name was not disclosed in the summary, discussed how AI services collect user data, often without clear limits on retention or reuse. He emphasized that many popular AI tools are not designed with privacy first. The “one thing” that worries him most is the fundamental lack of transparency and control that users have over their own data once it enters an AI system. Without strong regulation or technical safeguards, companies can use your inputs to improve their models, share them with third parties, or retain them indefinitely.

Proton itself has been adding AI features to its encrypted services, but always with end-to-end encryption and a commitment not to train models on user data without explicit consent.

Why it matters

For the average person, the privacy risks of AI are not always obvious. You might think that a conversation with a chatbot is ephemeral, but many platforms log every prompt and response. This data can include personal anecdotes, confidential work information, or even passwords accidentally pasted. Researchers have shown that some AI models can be tricked into reproducing private training data.

Beyond individual usage, the broader issue is the power imbalance between users and the companies running the AI. Users often have no way to know what happens to their data after they click “send.” The CEO’s concern reflects a growing unease among privacy advocates: that AI adoption is racing ahead without adequate guardrails.

What readers can do

You do not have to stop using AI tools to protect your privacy. The key is to be selective and deliberate. Here are practical steps you can take, many of which align with the advice from Proton’s CEO:

  1. Use privacy-focused alternatives. Services like Proton’s own AI features (available in ProtonMail and Proton Drive) are designed to process data without storing it in plaintext. Other options include locally running models such as Llama or Mistral, which never send data to a remote server.

  2. Avoid sharing sensitive information. Treat any conversation with a cloud-based AI as potentially public. Do not paste passwords, financial details, medical records, or anything you would not want a stranger to read.

  3. Review app permissions and data policies. Before signing up for an AI tool, read its privacy policy—or at least look for sections on data retention, sharing, and model training. Some services let you opt out of training, but you have to find the setting.

  4. Use a VPN and encrypted browser. ProtonVPN (or another reputable VPN) can mask your IP address, making it harder for AI platforms to link usage to your identity. Pair it with a privacy-oriented browser like Brave or Firefox with tracking protection.

  5. Audit your current AI usage. Check which AI tools you have accounts with, and delete any data or conversations you no longer need. Many platforms allow you to export and delete your history.

  6. Consider self-hosted or local alternatives. Tools like Ollama or GPT4All let you run AI models on your own computer. You lose some convenience, but your data never leaves your device.

The CEO’s overarching message was that privacy is possible if you choose the right tools and stay aware of the trade-offs. The thing that keeps him up at night is not that AI is inherently dangerous, but that most users do not realize how much control they have given away.

Sources

  • “Privacy in the AI era is possible, says Proton’s CEO, but one thing keeps him up at night.” Spiceworks, June 4, 2026. (Summary via Google News)
  • Proton’s official blog and documentation on AI features and data handling.

Note: The interview details are based on a news summary. For the full interview and exact quotes, consult the original Spiceworks article.