Originally researched at blog.useinflux.co: LinkedIn engagement pod rules
If you write software, LinkedIn is a tempting automation target. The feed is repetitive, commenting takes time, and your growth backlog probably has a ticket that says "schedule posts and auto-engage". Before anyone writes a line of code or installs a browser extension, it is worth reading what LinkedIn actually says. The rules are more specific than most people assume, and they draw a clear line between permissioned API access and scripts that act as if they were you.
This piece walks through the primary sources, quotes the relevant clauses, and ends with a short decision table you can use in a planning meeting. Every LinkedIn claim links to the page it comes from. I am not a lawyer, and nothing here is legal advice; it is a close reading of public pages, re-checked on 8 November 2026.
1. Start with Section 8.2 of the User Agreement
LinkedIn's User Agreement (effective 3 November 2025) has a "Don'ts" list in Section 8.2. Three clauses matter most for anyone building tooling (LinkedIn User Agreement):
- You agree not to "develop, support or use software, devices, scripts, robots or any other means or processes (such as crawlers, browser plugins and add-ons or any other technology) to scrape or copy the Services, including profiles and other data from the Services".
- You agree not to "use bots or other unauthorized automated methods to access the Services, add or download contacts, send or redirect messages, create, comment on, like, share, or re-share posts, or otherwise drive inauthentic engagement".
- You agree not to "override any security feature or bypass or circumvent any access controls or use limits of the Services".
Read the second clause carefully as an engineer would. The list of verbs is long: access, add contacts, download contacts, send messages, create posts, comment, like, share, re-share. The qualifier is "unauthorized automated methods". So the question for any tool is not "is this automation?" but "is this automation LinkedIn has authorised?"
The same section's first item also rules out using another person's account, giving "sharing log-in credentials or copying cookies" as examples. Any tool that asks for your password or session cookie deserves a hard look against that line.
2. The Help Centre names the tool categories
LinkedIn's Help Centre page on prohibited software makes the categories explicit. It says LinkedIn does not "permit the use of any third party software, including 'crawlers', bots, browser plug-ins, or browser extensions that scrape, modify the appearance of, or automate activity on LinkedIn's website" (LinkedIn Help: Prohibited software and extensions).
It also spells out the consequence: members who use such tools "risk having their accounts restricted or shut down", and the tools themselves "may become non-operational without notice". For a product team, that second line is a reliability warning as much as a policy one. A feature that depends on an unofficial extension can break the day LinkedIn changes its defences, and the page says LinkedIn is "constantly working to improve" them.
3. Comment automation has its own 2026 statement
On 13 March 2026, LinkedIn's Pressroom published a post specifically about authentic content and conversations (LinkedIn Pressroom). Its definition is useful because it is precise:
"Automated comments are comments posted to LinkedIn using a browser extension, script, or third party tool. They don't include human action or clicking the 'comment' button, and they aren't allowed on LinkedIn."
The post lists two things LinkedIn may do if it detects automation tools: not show detected automated comments in the "Most Relevant" comments (and sometimes not show them outside the commenter's network), and restrict a member's ability to use LinkedIn. It also says LinkedIn has been reaching out to members whose actions showed signs of engagement pods or automation tools, warning of possible account restrictions and removal from LinkedIn programmes, including Top Voices.
Two implications for builders:
- The test LinkedIn describes is human action. An AI tool that drafts a comment for a person to read, edit and post by hand is a different thing from a script that posts it. The first keeps a human click in the loop; the second does not.
- Even when an automated comment is not removed, LinkedIn says it may be pushed out of "Most Relevant". So the expected reach of automated engagement is low before you even count the account risk.
4. Coordinated engagement is covered too, automated or not
Some teams avoid scripts and instead coordinate people: a shared channel where colleagues are asked to like and comment within the first hour. LinkedIn's Professional Community Policies address this directly: "Don't do things to artificially increase engagement with your content. Respond authentically to others' content and don't agree with others ahead of time to like or re-share each other's content" (LinkedIn Professional Community Policies).
The Pressroom post calls an engagement pod "a coordinated group of people that like, comment and share posts to attempt to boost visibility and virality" and says this "is not allowed on LinkedIn" (LinkedIn Pressroom). So the policy concern is the coordination and the artificial boost, not only the code. Telling colleagues a post is live is ordinary; agreeing in advance to engage with each other's posts is what the policy text describes.
5. There is an official, permissioned route for posting
None of this means LinkedIn bans software from posting. LinkedIn's developer documentation on Microsoft Learn describes a "Share on LinkedIn" product. It uses the OAuth 2.0 flow, and the guide says the w_member_social permission is "required to create a LinkedIn post on behalf of the authenticated member" (Microsoft Learn: Share on LinkedIn).
That is the distinction the User Agreement's word "unauthorized" points at. A scheduler that asks the member to sign in through LinkedIn's OAuth screen, receives a scoped token, and publishes a post the member wrote is using a documented route. A script that logs in with a stored password or copied cookie and clicks around the website is not.
One practical note: the Share on LinkedIn guide on Microsoft Learn carries a 2021 date, and LinkedIn has versioned its APIs since, so check the current API reference and your app's approved products before you build.
A decision table for your next planning meeting
Use this as a first filter, then read the sources yourself.
| What the tool does | How it does it | What LinkedIn's pages say |
|---|---|---|
| Publishes a post the founder wrote | OAuth sign-in, scoped token, documented API | Documented route (Share on LinkedIn, w_member_social) |
| Drafts comments for a human to review and post by hand | AI or templates, human clicks "comment" | Human action is the test in the Pressroom definition |
| Posts comments automatically | Browser extension, script or third party tool | "Aren't allowed on LinkedIn" (Pressroom, 13 March 2026) |
| Likes, shares or adds contacts in bulk | Bot or session automation | Listed in User Agreement 8.2 as prohibited when unauthorised |
| Scrapes profiles into a CRM | Crawler or extension | Listed in User Agreement 8.2 and the Help Centre page |
| Coordinates colleagues to engage in the first hour | Group chat or rota | Professional Community Policies: don't agree ahead of time to like or re-share |
What I would automate instead
If the goal is founder time saved, the safest wins are upstream of LinkedIn itself:
- Capture: record a 20 minute voice note a week and transcribe it into a topic backlog.
- Drafting: generate first drafts from the founder's own notes, then edit them in the founder's voice.
- Scheduling: publish through an OAuth-based tool, or simply set a reminder and post by hand.
- Measurement: export your own analytics and track a small set of outcomes, such as replies from target roles and "where did you hear about us?" answers on discovery calls.
- Engagement: keep commenting human. Use software to surface posts worth replying to, not to reply.
That split keeps every action LinkedIn's pages focus on (posting, commenting, liking, connecting) either behind a documented permission or behind a human click.
Sources (all re-fetched 8 November 2026)
- LinkedIn User Agreement, Section 8.2: https://www.linkedin.com/legal/user-agreement
- LinkedIn Help, Prohibited software and extensions: https://www.linkedin.com/help/linkedin/answer/a1341387
- LinkedIn Pressroom, What We're Doing to Support Authentic Content and Conversations on LinkedIn (13 March 2026): https://news.linkedin.com/2026/authentic-content-and-conversations
- LinkedIn Professional Community Policies: https://www.linkedin.com/legal/professional-community-policies
- Microsoft Learn, Share on LinkedIn: https://learn.microsoft.com/en-us/linkedin/consumer/integrations/self-serve/share-on-linkedin
Originally researched at blog.useinflux.co. The full breakdown of LinkedIn's engagement pod rules is here: https://blog.useinflux.co/articles/linkedin-engagement-pods-banned
Disclosure: I am the founder of Influx (https://useinflux.co), which works on LinkedIn authority and distribution for founder-led B2B companies.
Top comments (0)