How to Use AI Without Sacrificing Your Privacy – Advice from Proton’s CEO

How to Use AI Without Sacrificing Your Privacy – Advice from Proton’s CEO

If you’ve ever typed a sensitive question into a chatbot or copied a private document into an AI summariser, you’re not alone. AI tools have become everyday helpers. But the trade‑off between convenience and privacy is one that many users don’t realise they’re making. In a recent interview with Spiceworks, Proton’s CEO Andy Yen argued that privacy in the AI era is achievable – yet he admitted that one issue keeps him up at night.

What happened

In the interview, Yen said that the core privacy problem with today’s AI tools is the way they collect and use personal data. Many AI services store your conversations, use them to improve their models, and share data with third parties – often without clear explanations. Yen pointed out that the lack of visibility into how your data flows makes it hard for users to trust these systems. He warned that people are increasingly locked into platforms that treat personal information as a resource for training, rather than something to be protected.

Proton, the company behind Proton Mail, VPN, and Drive, has built its reputation on end‑to‑end encryption and a zero‑access architecture. Yen said Proton is working on bringing similar principles to AI, but he emphasised that the industry as a whole still has a long way to go.

Why it matters

For everyday users, the implications are immediate. When you ask an AI assistant for help with a personal letter, a work problem, or even health advice, that input can become part of the model’s training data. Companies differ in how they handle this – some let you opt out, others default to inclusion. And policies change often.

Even if a service promises not to use your data for training, there is still the risk of data breaches or legal requests for stored logs. With cloud‑based AI, you often don’t know where your data resides or who has access to it. In a world where data is increasingly valuable, this lack of control can feel unsettling.

What readers can do

There is no single perfect solution, but you can reduce your exposure with a few deliberate choices:

• Check the privacy settings of any AI tool you use. Most major platforms – ChatGPT, Google Gemini, Microsoft Copilot – allow you to turn off chat history or disable model training. This is often buried in settings menus, but it’s worth the time.
• Use services that don’t store your data by default. Some AI tools are built with privacy in mind. They may run locally on your device or use end‑to‑end encryption. Proton has hinted at AI‑powered features that follow the same zero‑access model, but as of mid‑2026, it remains to be seen how fully the company will deliver.
• Limit what you share. Even if you trust the service, consider how much personal detail you include in prompts. Names, addresses, and financial information are best kept out of AI conversations altogether.
• Read the privacy policy – or at least the summary. Many services now provide short “data use” explanations when you first sign up. A five‑minute read can tell you whether your inputs will be saved, for how long, and with whom they may be shared.
• Prefer on‑device AI when possible. Tools that process everything on your phone or laptop (like Apple’s on‑device models or some open‑source alternatives) never send your data to a server. This avoids many of the data‑collection risks, though it may require more powerful hardware.

Sources

The interview with Proton’s CEO was originally published by Spiceworks. You can read the full article here: Privacy in the AI era is possible, says Proton’s CEO, but one thing keeps him up at night.