Can You Trust Perplexity?

Perplexity is a great product. Its search experience is fast, useful, and often better than traditional search when you want a direct answer with sources.

But that is not the question.

The question is: can you trust Perplexity with your private data and your confidential work?

That question matters more now because Perplexity has just launched something called Personal Computer. On paper, it is a smart idea. In practice, the name hides the most important part of the tradeoff.

What Perplexity’s Personal Computer is

Perplexity says Personal Computer runs on a dedicated Mac mini that can operate 24/7, connected to your local apps and Perplexity’s secure servers. The company describes it as a digital proxy for you, one that works constantly on your behalf and lets you orchestrate your tools, tasks, and files from any device, anywhere. Perplexity also says the system runs in a secure environment, sensitive actions require approval, every session has an audit trail, and there is a kill switch. The product is currently on a waitlist.

That sounds impressive. It probably will be impressive.

But it is also where the language starts to get slippery.

The problem is the word “personal”

Merriam-Webster defines personal as “of, relating to, or affecting a particular person,” and also as something “private” or “intended for private use or use by one person.” Merriam-Webster defines private as something “intended for or restricted to the use of a particular person” and “not known publicly.”

That is how normal people understand the word too.

A personal computer is supposed to feel like your machine. Something you own. Something under your control. Something you can use even offline. Something that gives you privacy because it belongs to you and lives with you.

That is not really what Perplexity is selling here.

Why the branding is smart, and why it is confusing

Perplexity’s marketing is clever because it shows a Mac mini-like setup and wraps it in the language of ownership. “Personal Computer” sounds like the machine is yours in the old-fashioned sense.

But Perplexity’s own description says it is connected to Perplexity’s secure servers and designed to be controlled from any device, anywhere. In other words, this is not just “your Mac.” It is your workflow, merged with Perplexity’s cloud system and Perplexity’s control layer.

That may be convenient. It may even be extremely useful.

It is just not what many people mean when they hear the word personal.

The real issue: who controls the data

This is where the privacy story gets much less comfortable.

Perplexity’s own help center says that for Free, Pro, and Max users, AI Data Retention is enabled by default. It also says that if the toggle is enabled, your data is being collected for AI training. Users can opt out, but Perplexity says that opt-outs apply only to future data, and that previously collected training data cannot be deleted or removed. Perplexity also says data may still be processed for service operation, legal compliance, and product improvement.

That is the part professionals should focus on.

If you are handling confidential files, your default should not be “I hope the settings are right.” Your default should be “this never leaves my machine.”

Enterprise gets a different promise

Perplexity’s enterprise pages make a much stronger privacy promise. The company says it never trains its LLMs on enterprise customer data, offers configurable file retention, and provides audit logs and access controls. Its help pages also say enterprise query information is never used for training.

That contrast matters.

It means Perplexity itself is acknowledging two different privacy standards:

• one for enterprise customers, where “no training” is guaranteed

• one for normal users, where training is enabled by default unless you change the setting yourself

So when a consumer-facing product is called Personal Computer, you should not automatically hear “private by default.” You should ask which privacy model applies.

This is why “personal” should mean ownership, not delegation

A real personal computer is still the simpler thing:

It is the computer you can touch. The one you own. The one that keeps working when the internet is down. The one where your files stay local unless you explicitly move them.

Perplexity’s Personal Computer may be powerful, but it is still a system where a big tech company sits in the middle of your work. And if that work includes confidential files, that middle layer is the whole problem.

Perplexity is not (reportedly) currently leaning into ads

One part of the usual cloud-AI fear is advertising. On that point, Perplexity has actually moved away from ads, saying they could undermine trust because users might become “suspicious of everything.” That is a smart move, and it deserves to be said clearly.

But it also proves the bigger point: even Perplexity knows that trust is fragile when a platform sits between you and the answer.

So can you trust Perplexity?

For general web search and lightweight tasks, maybe.

For private data and confidential professional work, I would be much more cautious.

Because a personal computer should first mean ownership, locality, and privacy. It should not mean “your tasks, routed through a company’s servers, under a policy model that trains on consumer data by default unless you opt out.”

That is why privacy is freedom.

If you want AI help with your files, the better model is keeping that work on your own machine. On Mac, that is where Fenn fits.

Fenn is Private AI that finds any file on your Mac. It searches inside your files and helps you work with them locally, without turning “personal” into a cloud marketing word.