Source telegram blog remains the main reference point for users and Telegram Mini App developers following this update.
Telegram has introduced more than ten new features that reshape how users work with AI bots and manage chats. The headline: AI assistants and automated tools—built as Telegram bots—can now be mentioned in any private or group chat, even if they’re not chat members. To get a bot’s attention, simply tag it with its @username in a message. The bot will only receive the message in which it’s tagged and any direct replies. It does not see chat member lists or other parts of the conversation. This creates a clear privacy boundary, while making bot-powered responses instantly accessible.
Bot creators and admins will also see a big shift: bots can now talk to other bots. This unlocks new workflows, like bots handling tasks for each other, automated inbox management, and more advanced chat operations for teams and community managers. According to the Source telegram blog, you can even link a bot to your user profile for automated responses, choosing exactly which chats the bot may access or ignore. These API updates give both users and developers tighter control over automation and personalization inside Telegram.
Source telegram blog: How Guest AI Bots Work in Telegram Chats
Guest AI Bots now let anyone access AI services from any group or personal chat, just by mentioning a bot’s @username—even if the bot hasn’t been added to the chat. When a bot is tagged, it sees only that single message and any replies to it, and can reply right there in-line. Bots do not have access to previous messages, participants’ identities, or the broader chat history. User data and conversation details are sharply limited.
The process is simple: mention the bot’s @username in your message. If that bot supports guest mode, it’ll respond publicly in that thread. Replies are streamed as soon as they’re generated, so you don’t have to wait for the AI to finish its answer to start reading. Developers have to opt a bot in to guest mode, and Telegram’s developer platform provides guidance on getting started.
TON Drop Hub take: Every chat—personal or group—can now tap into AI tools without cluttering up the member list with bots. Privacy is preserved by design: bots only see what they need for specific requests. Before tagging a bot, make sure you’re using the real @username and not a copycat, as phishing and impersonation are ongoing risks.
Managing Bot Permissions and User Privacy
Telegram now gives users granular control over what bots can see and do. Bots only receive the message they’ve been tagged in (and direct replies), not the chat history or participant list. For example, if you tag a bot with @username in a chat, it’s blind to all previous messages and doesn’t know who else is in the conversation. This sharply cuts down the data any single bot can collect.
Users linking bots to their own profiles can decide exactly which chats a bot can work in. You can exclude existing contact chats or focus bot automation only on new conversations. There’s no all-access pass—bots stay within boundaries you set. Change these options by going to Settings > Chat Automation. Keeping the control panel accessible reduces the chances of accidental data exposure or bots reaching into unintended chats.
For developers, these updates are significant for privacy compliance and transparency. Now, user data handling is much clearer and more restricted. Bots only have what’s necessary to function, creating a more trustworthy foundation for chat-based utilities.
TON Drop Hub take: For active campaigns or airdrop quests using bots, restricted permissions mean lower data risk. Developers need to double-check how their automation works to fit these new limits—but privacy as the house default is an upgrade for everyone.
Creating and Sharing Custom AI Styles in Telegram
Telegram now offers users a way to generate their own AI writing styles through the text editor. Simply go to the ‘Style’ tab and tap ‘Create’ to set up a custom prompt. Each new style you make produces a link that anyone can preview and add. This is especially useful for admins or creators looking to push consistent templates or meme formats across a channel or group.
There isn’t a public record of moderation for these AI styles yet, so anyone can create and share whatever they like. If you’re adding a style from a link, check the preview and sample some generated text before integrating it. Telegram hasn’t given hard limits on the number of styles you can manage, and it’s unclear how edits or updates are handled after sharing.
TON Drop Hub take: User-generated text styles should speed up content workflows across larger groups and projects—but users need to be careful about where these links come from and evaluate the results, since there’s no curation or moderation at rollout.
Telegram’s most recent update gives bots and AI a stronger presence in daily chats. Bots can now be summoned as needed, interact with each other, and operate without deep access to personal or chat data. Real-time bot replies and hands-on permission controls keep automations powerful but contained. Connecting bots to your profile now puts you in charge of exactly which chats they touch.
TON Drop Hub take: For builders and advanced users—including those customizing wallets, running DeFi bots, or managing community quests—these changes mean more secure and flexible automation. Less friction, more oversight.
For more ecosystem coverage, see Latest TON news. Source reference: original source.