Is this Bluesky account a bot?
Look for what an account declares: an Automation self-label, an attributed bot label or a verified bot handle. The absence of these signals does not establish whether the account is automated.
By The Bluesky.bot team
- Published:
- Updated:
- Reviewed:
Check the public declarations
Examples:demo.bluesky.botbsky.app
The checker reports present, absent or unknown for each source. It does not decide whether a person or a program operates the account. A biography is displayed as self-description; its wording is not a detection rule.
The Automation label
Bluesky’s Automation marker comes from the account’s own bot self-label on its profile record. The AT Protocol bot tutorial recommends identifying bot accounts this way as a best practice. The account controls this declaration; it is not a Bluesky identity verification badge.
The measured share uses the directory’s last recorded label observations for publicly listed bots whose handle is live. It does not describe all accounts on Bluesky. A missing or suppressed denominator is not replaced by a guess.
Labels have a source
A label names the DID of the labeler that issued it, its subject and its value. Revocations and expiration matter. Our checker uses the label inspector with a fixed scope: only Bluesky.bot’s labeler, when available. It shows attributed labels such as verified-bot, certified-bot and official-bot, plus creator-declared descriptive values. An AI-generated descriptive label alone does not establish automation. Other labelers may hold different information.
To understand subscriptions and viewing choices, read how Bluesky works. A label is a statement by its source, not a consensus of the network.
What a bot handle establishes
A live name.bluesky.bot handle links control of that name to the account’s DID. A public listing can add a category, an operator line and a usage card. Handle control does not endorse every post or guarantee good conduct. The verification guide explains the identity question in more detail.
Why we show declarations instead of a score
A regular posting time can come from a person using a scheduler; a bio can be humorous or outdated. Behavioural models depend on their training data and assumptions, and a probability is not proof about an individual account. We do not infer automation from wording, follows, replies, posting rhythm or AI-text detection. The result remains a list of verifiable declarations with their sources and uncertainty.
Handling unwanted interactions
Use Bluesky’s mute and block controls to manage unwanted interactions, and report conduct that breaks its Community Guidelines. You do not need to establish automation to decide what you want to see. Choose moderation lists or labeler subscriptions according to their stated purpose and your viewing preferences.
For creators
Set an accurate Automation declaration, choose a recognisable handle, and make your operator and usage instructions clear. See the Automation label guide and the bot handle guide. Keep those declarations current when the account’s purpose changes.