Is Plaud AI safe? Where your recordings go, and who hears them

The Morse teamPublished November 1, 20267 min read

As of November 2026, Plaud AI is reasonably safe with your recordings by its own account, but it isn't on-device AI: Plaud's trust center says "All data is transferred to the cloud for processing," and transcripts and summaries are made by GPT, Claude, or Gemini, even when you've left cloud storage off (Plaud trust center). Since June 2026, new users also have transcription switched on for every recording by default, and whether the people you recorded agreed is, in Plaud's words, "your responsibility" (Plaud release notes, Plaud privacy policy).

We read Plaud's privacy policy, its AI Data Usage Transparency Policy (updated May 20, 2026), its trust center, its release notes, a Plaud support article, and the Plaud app's App Store listing on November 1, 2026, and took Plaud at its word; we ran no tests. Plaud earns credit for its training rule. Its privacy policy says it will never use your recordings, transcripts, or files to train its own or anyone else's AI "unless we have obtained your explicit and informed consent in advance," and it binds its AI providers not to train either (Plaud).

Plaud privacy settings at a glance

Personal accounts, read November 1, 2026.

Does Plaud process recordings on the device or in the cloud?

In the cloud. Plaud's trust center answers its own question on this directly: "All data is transferred to the cloud for processing" (Plaud). The device itself only records. Plaud says recordings "will be deleted from the device once they are transferred" to your phone, and that only the Plaud account the device is bound to can decrypt them.

Plaud's transparency policy can read like the opposite. It says "By default, your data is stored locally on your device and is not uploaded to any server" (Plaud). That sentence is about storage. With Private Cloud Sync off, the trust center says audio and transcripts go to the cloud to be processed and are deleted from it "once it is transferred back to the phone." Turn sync on, and your files stay on Plaud's servers until you delete them.

The phone app is the only way to keep sync off. Plaud's privacy policy says that for the Web App and Desktop App, "cloud synchronization is required in order to provide the Services" (Plaud).

Which AI models hear your Plaud recordings?

GPT, Claude, or Gemini. Plaud's trust center names "GPT, Claude, and Gemini" as the models it uses, says users in Europe get "EU-hosted, enterprise-grade subprocessors," and says "the rest of the world will rely on US-hosted LLMs" (Plaud). For transcription, it says it uses "multiple enterprise-grade models," including its own fine-tuned ones.

Storage runs on Amazon Web Services. Plaud lists data centers in "US AWS-West, EU AWS (Frankfurt), Japan AWS, and Singapore AWS," with your data kept in the region nearest you.

Finding those names takes some looking. Plaud's privacy policy says the identity of "our third-party AI model providers" is in its AI Data Usage Transparency Policy (Plaud). The version of that policy we read, updated May 20, 2026, names no provider; it says only "AI service providers" (Plaud). The trust center is where the names are.

How long do Plaud's AI providers keep your audio?

Plaud describes it three ways. The trust center says "All LLMs used in Plaud have zero training and zero data retention policies, enforced contractually via DPAs" (Plaud). A Plaud support article updated October 8, 2026 says the data "is deleted immediately after processing" (Plaud support). The privacy policy is looser: provider agreements require that data "is retained only for a limited period as necessary to provide the service and meet security and compliance requirements" (Plaud). Our explainer on zero data retention covers what that promise usually allows.

The same support article answers a question healthcare users ask. Plaud "does not automatically redact patient-identifiable information" such as names, dates of birth, or medical record numbers spoken in a recording before an AI model processes it. Plaud's own advice is to say "Patient A" instead.

Does Plaud transcribe recordings automatically?

Yes, for new users, since June 15, 2026. Plaud's release notes say "AutoFlow is now on by default," so "recordings captured will be transcribed instantly," and that you can turn it off in the welcome tour or in Settings (Plaud). The App Store notes for version 3.25.0 say the same of Auto-Transcribe (App Store).

That default sits awkwardly with the transparency policy, which still says transcription and summaries "are initiated only at your request" and that Plaud does "not process your data in the background without your action" (Plaud). If you'd rather choose which recordings leave your phone, switch AutoFlow off before your first recording syncs.

Plaud Desktop goes further on a computer. Its privacy policy says the desktop app "may also offer an automatic recording feature" that starts when it detects a meeting app running, and uploads those recordings to your cloud (Plaud).

Can Plaud employees access your recordings?

In limited cases, by Plaud's trust center. It says personnel "do not routinely access users' data," then lists when they may: when you ask for technical support that needs your data, to diagnose system errors or security incidents, and to comply with the law (Plaud). Access is logged and needs internal approval.

The transparency policy states it more absolutely: "no third parties, including our employees, have direct access to your data" (Plaud). Read the two together and the safe assumption is that a support ticket can open your recording to a Plaud engineer. Courts can reach it too. The privacy policy says Plaud may disclose data "in response to a valid legal process, such as a court order, subpoena," and synced recordings sit on its servers for as long as your account is active. Our guide to AI chats in court shows how that has played out for chatbots.

Is it legal to record people with a Plaud?

It depends on where you are, and Plaud puts the duty on you. Its privacy policy says "complying with the recording-consent laws that apply where you record, is your responsibility," and the footer of Plaud's site asks you to "let others know and get their okay" before you record (Plaud). This isn't legal advice.

US federal law lets you record a conversation you're part of, since it isn't unlawful where "such person is a party to the communication" (18 U.S.C. § 2511(2)(d)). Some states require everyone's consent. California makes it a crime to record "a confidential communication" without "the consent of all parties," with a fine of up to $2,500 per violation (California Penal Code § 632). A pin on your shirt is easy to miss, so say you're recording.

How Morse keeps your notes off a transcription service

Every question above begins with audio. A recording of everyone in the room goes to Plaud's cloud, then to GPT, Claude, or Gemini, and the people on it may never have agreed to any of that. Morse is an AI chat for the part afterwards, when you type up what was decided or paste in a few lines of notes and ask for a summary, a follow-up, or a plan, and only the words you choose go in.

• E2EE, for notes about clients, patients, or anything no one else should read. Your browser checks the secure enclave's proof that it's genuine, then encrypts your message so only that enclave can read it. Morse's servers relay text they can't read. This rests on the enclave hardware and attestation working as designed. E2EE chats come with Pro, $19 a month.
• Private, for everyday notes. New chats start in Private. Morse has zero data retention agreements with its inference, search, and other providers, so they keep nothing, and where a provider confirms that on each reply, Morse checks it.
• Anonymous, for GPT, Claude, or Gemini without your name attached. Our proxy strips your account, email, and IP address, so the provider sees the message without knowing it's you. It may keep it for a limited time to check for abuse under its own policy, so leave names out of what you paste.

In every mode, your chat history is stored in your browser, encrypted, not on our servers, and it locks when you sign out. Morse doesn't train AI on your content, sell your data, or show ads (Morse privacy policy). Because we don't store chats, we usually have very little to give authorities. Clearing your browser or deleting your account deletes your chats, and Morse can't recover them, so keep a copy of anything you need.

So is Plaud safe enough for you?

• Cloud storage: use the phone app with Private Cloud Sync off, and know that transcription still runs in Plaud's cloud. On Morse, your chat history stays in your browser.
• Automatic transcription: turn AutoFlow off and transcribe only the recordings you choose.
• Other people's privacy: tell everyone before you record, and don't name speakers you haven't asked.
• Health and client notes: ask Plaud for a BAA before recording patients, and don't say identifiers out loud. For notes typed afterwards, use an E2EE chat on Morse.

Read whether Otter AI is private if your recordings come from meeting software instead. To try Morse, send a chat from the Morse home page without an account, or start free with no card needed.