---
title: "Can People Tell If a DM Is Automated? 6 Tells Leads Notice"
slug: can-people-tell-when-a-dm-is-automated
author: "Leonardo Maldonado"
category: "AI Setter Operations"
articleType: how-to-guide
tags: ["automated DMs","AI disclosure","Instagram DMs","bots"]
publishedAt: 2026-11-15T16:00:00.000Z
updatedAt: 2026-11-15T16:00:00.000Z
canonical: https://setluca.com/blog/can-people-tell-when-a-dm-is-automated
---# Can People Tell If a DM Is Automated? 6 Tells Leads Notice

> Can people tell if a DM is automated? Often, yes, and the same tells show you how to tell if you are talking to a bot. Look for instant replies at odd hours, identical openers, answers that ignore what you wrote, stiff wording, wrong details and dodged questions. For a business, the goal is helpful, honest, accurate replies, plus a plain yes when a lead asks if it's AI.

## Key takeaways

- People notice automation (software that sends replies for you) through behavior, not vocabulary. Timing, sameness and missed details give it away faster than any single word.
- If you're on the receiving end, the quickest test is to ask. On Telegram, a bot account also shows a "bot" label in the chat.
- Meta recommends telling people when a chat is automated, and California and the EU push the same way.
- Fix the tells by writing replies to what the lead said, keeping facts accurate and pacing replies like a person.
- Never deny AI. If a lead asks "is this a bot?", tell them yes, then add something useful.

---

Can people tell if a DM is automated? Often they can, and if you're wondering how to tell if you are talking to a bot yourself, both questions share an answer. Automated DMs give themselves away through timing, sameness and replies that miss what was said. This guide covers those tells, the disclosure rules (the rules on telling people a chat is automated), and what to reply when a lead asks.

## Can people tell if a DM is automated?

Yes, leads can often tell a DM is automated, and they usually notice behavior before wording. A reply that arrives seconds after a 3am message, or one that skips the question someone asked, reads as automated. Meta says there are more than a billion active business threads a day, so most people have seen automated replies up close. A fast, specific, accurate reply rarely bothers anyone. A generic reply followed by a dodged question does.

Our post on [how Instagram detects automation](/blog/how-instagram-detects-automation) covers the platform's side. For how business inboxes work, see [Instagram direct messages for business](/blog/instagram-direct-messages-for-business). This one covers the human side.

## What tells do people notice in an automated DM?

People notice six tells in automated DMs. Each one is a behavior rather than a word choice, so each one has a fix you can check in your own inbox. The last two do the most damage, because they're accuracy and honesty problems, not style.

1. **Instant replies at 3am.** A full, polished answer seconds after a late-night message looks like nobody typed it. Fix: pace replies like a person.
2. **Identical openers.** Two friends compare screenshots and see the same first line. Fix: open with something from the lead's own message.
3. **Ignoring what they said.** The lead mentions a knee injury, and the reply talks about "your fitness goals." Fix: answer the detail first.
4. **Over-formal tone.** "Thank you for reaching out regarding our program" from a coach who writes in lowercase. Fix: train on your real sent replies.
5. **Wrong details.** A price you don't charge or a slot that's already booked. Fix: check every price, offer and date before it goes out.
6. **No answer to a direct question.** The lead asks "how much is it?" and gets another question back. Fix: answer first, then ask one question.

## How to tell if you are talking to a bot

Ask first. A plain "Am I talking to a person or AI?" is the fastest test, and an honest business will say yes if AI is replying. Then check two more things:

- **Look for a label.** Telegram's documentation says bots show a "bot" label in the chat instead of a "last seen" status.
- **Ask something off-script.** A question that mixes two of your own details tests whether anything actually read your messages.

No label doesn't prove a person is typing, because a person can draft with AI and send it themselves. Plenty of coaches already draft that way, as our guide to [ChatGPT for coaches](/blog/chatgpt-for-coaches) describes. So asking is still the most reliable check.

## Why is "can they tell" the wrong goal?

"Can they tell?" is the wrong goal because it aims at hiding, and hiding fails the first time someone asks. The goal that holds up is a reply that's useful, honest and accurate, sent at a sensible pace. Chasing "undetectable" pushes you toward fake typos or invented delays, and those small lies become a big one when a lead asks "is this a bot?"

Write replies the way you write, keep the facts right, answer what was asked, and say yes when someone asks about AI. Our comparison of an [AI setter vs a human setter](/blog/ai-setter-vs-human-setter) covers where each one earns its place.

## What do the disclosure rules say about automated DMs?

The short version: tell people a chat is automated, and never deny it when asked. Meta's messaging policy requires disclosure where the law requires it. California's Business and Professions Code Section 17941 bans bots that mislead about their identity to drive a sale. Article 50 of the EU AI Act applies from August 2, 2026. This is general information, not legal advice. See [what Meta allows for DM automation](/blog/what-meta-allows-for-dm-automation) and [is Instagram DM automation safe](/blog/is-instagram-dm-automation-safe) for the platform side.

## How does SetLuca write and pace replies?

SetLuca replies to people who wrote to you first, within minutes at any hour, paced like a person, inside each platform's reply window. It never messages anyone cold.

It learns from replies you paste in at setup and from your live conversations, and the writing-style profile is visible and editable. Every draft gets a second-pass check for made-up prices, offers or availability before it goes out. Our guide on [how to train AI to write like you](/blog/how-to-train-ai-to-reply-in-your-voice) shows the method.

## What should you say when a lead asks if it's a bot?

Say yes, plainly, then keep helping. A line like "Yes, replies here are written with AI, set up from the way I write, and I take every call myself" works. A dodge usually ends the conversation. For harder pushback, our guide to [objection handling in DMs](/blog/objection-handling-in-dms) has more.

## A worked example: one late-night reel

Here's how one late-night reel plays out for Rafa, a fictional run coach (illustrative only). A reel posted at 9pm brings in DMs all night.

Rafa's old keyword flow, a tool that fires a set message when a DM arrives, sent one opener to every new DM within seconds: "Hey. Thanks for reaching out. What are your running goals?" One lead had asked whether the plan suited a calf tear. At 2:40am she wrote "wait is this a bot lol," got the opener again, and left. That's four of the six tells at once. The honest version:

### What does the honest version look like?

> **Lead:** "I tore my calf last spring, is this plan OK for me?" **Reply:** "Calf tears are common with runners. The plan builds mileage slowly, but your physio should clear you first. Are you running at all right now?" **Lead:** "wait is this a bot lol" **Reply:** "Yes, it's AI, set up from the way Rafa writes. He reads every thread and takes every call. If you'd rather wait for him, he'll pick this up in the morning."

## Where does a human still read DMs better?

One honest limit: a great human setter still reads hard, emotional conversations better than AI does. A lead who's anxious, grieving or burned by a past coach needs someone who can slow down. SetLuca books the call, and you take it and close.

## Should you worry whether leads can tell?

Worry less about detection and more about the reply. If a DM is useful, honest and accurate, a lead who knows AI helped usually doesn't mind. Disclose up front when the law or the stakes call for it:

| Situation | What to do |
| --- | --- |
| Leads in California or the EU | Disclose up front in your profile and first reply |
| Leads elsewhere, high-trust offer | Disclose up front, as Meta recommends |
| Any lead who asks | Say yes, then answer the question |

So can people tell if a DM is automated? Often, and the honest answer to "how do I tell if I'm talking to a bot" is: ask, and expect a straight yes. Give your leads that same answer, every time. Then spend your effort on what they notice: the timing, the details and whether their question got answered.


## FAQ

### How can you tell if you are talking to a bot in DMs?

Ask directly whether you're talking to a person or AI, and expect an honest answer. Watch for instant replies at odd hours, identical openers, replies that ignore what you wrote and dodged questions. On Telegram, bot accounts show a "bot" label in the chat instead of an online status.

### Can people tell when an Instagram DM is automated?

Often, yes. People notice behavior more than wording: replies seconds after a late-night message, the same opener a friend received, or an answer that skips their question. Replies written to what the lead said, with accurate facts and sensible pacing, read far less like automation, but honesty still matters most.

### Do I have to tell leads my DMs are automated?

It depends on where your leads are. Meta requires disclosure where the law requires it and recommends it otherwise. California bans bots that mislead about their identity to drive a sale, and the EU AI Act requires disclosure from the first interaction. Never deny AI when asked.

### What should I say if a lead asks "is this a bot?"

Say yes, plainly, then keep helping. For example: "Yes, replies here are written with AI, set up from the way I write. I take every call myself." Offer a path to you if they'd prefer it. A dodge ends more conversations than a clear yes does.

### Does SetLuca pretend to be human?

No. SetLuca replies within minutes and paces messages like a person, but the aim is accurate replies, not hiding the AI. Write your own disclosure line, and ask SetLuca support how first replies work. If a lead asks whether they're talking to AI, tell them yes. Auto-send is on by default.


## Sources

1. [Meta for Developers, "Messenger Platform and IG Messaging API policy," 2026. Retrieved 2026-10-26.](https://developers.facebook.com/documentation/business-messaging/messenger-platform/policy)
2. [Meta for Developers, "Instagram API with Instagram Login: Send Messages," 2026. Retrieved 2026-10-26.](https://developers.facebook.com/documentation/instagram-platform/instagram-api-with-instagram-login/messaging-api)
3. [Telegram, "Bots: An introduction for developers." Retrieved 2026-10-26.](https://core.telegram.org/bots)
4. [Meta, "Be There for Every Customer With Meta Business Agent," 2026. Retrieved 2026-10-26.](https://about.fb.com/news/2026/06/meta-business-agent/)
5. [California Legislature, "Business and Professions Code Section 17941" (SB 1001), 2018. Retrieved 2026-10-26.](https://leginfo.legislature.ca.gov/faces/codes_displaySection.xhtml?lawCode=BPC&sectionNum=17941.)
6. [European Commission, "Transparency obligations under Article 50 of the AI Act," 2026. Retrieved 2026-10-26.](https://digital-strategy.ec.europa.eu/en/faqs/transparency-obligations-under-article-50-ai-act)
7. [NYC Department of Consumer and Worker Protection, "Mamdani Administration Proposes Ban on AI Customer Service Doom Loops," 2026. Retrieved 2026-10-26.](https://www.nyc.gov/site/dca/news/072-26/mamdani-administration-proposes-ban-ai-customer-service-doom-loops)
8. [SetLuca, "Pricing," last updated 2026-10-21. Retrieved 2026-10-26.](https://setluca.com/pricing)

---

Published by SetLuca.