What LinkedIn's User Agreement Actually Says, Verbatim
Directly from LinkedIn's User Agreement, confirmed via direct fetch of the live document: it prohibits developing, supporting, or using software, devices, scripts, robots, or other means or processes (including crawlers, browser plugins and add-ons, or any other technology) "to scrape or copy the Services." Separately, it prohibits using "bots or other unauthorized automated methods to... add or download contacts, send or redirect messages," and it prohibits bypassing the platform's security or access controls. These are not paraphrased summaries. They are the platform's own words, and they cover most of what "LinkedIn automation" sales tools actually do.
Why This Clause Exists, and Why It Is Written So Broadly
The broad language is not an accident or a loophole waiting to be found. It is written to cover scraping, bot-driven connection requests, automated messaging sequencers, and access-control bypass tools all under one policy, because any of them can be used to extract or send at a scale and speed a human operator working the platform manually cannot match. A tool marketed as "LinkedIn automation for sales teams" is, by the plain language of the agreement, describing exactly the category of activity this clause targets.
A Real-World Enforcement Example, Attributed Honestly
This risk is not hypothetical. Trade press coverage, not a primary company statement, reported that Artisan AI's own team and founder LinkedIn accounts faced restrictions or bans in late 2025, tied to suspected automation or spam violations. Treat this as directional, trade-press-sourced information rather than an audited enforcement record, since neither LinkedIn nor Artisan has published an official account of the incident that this research could locate. The detail worth sitting with regardless: this reportedly happened to a company whose own product sells automated outbound tooling, a notable data point on how seriously the underlying platform risk should be taken by anyone building an outbound motion around it.
What Counts as "Automation" in Practice
Not every LinkedIn activity by a sales rep crosses this line. A person manually sending individually written connection requests and messages, at a human pace, from their own account, is a fundamentally different activity than a browser extension auto-sending connection requests on a timer, or a sequencing tool posting messages on a schedule without a person actively present for each one. The clause targets the automated, unattended version of the activity, not the fact that a salesperson is using LinkedIn to prospect at all.
Why This Matters More for a Company Selling Outbound as a Service
If a vendor's entire top-of-funnel motion runs through LinkedIn automation tooling, that vendor is carrying a platform-level risk that could affect every client campaign running through the same infrastructure at once, not just one account. That is a structural question worth asking any vendor directly before you hand them your outbound: what happens to my campaign if your sending infrastructure gets flagged, not because of anything I did, but because of how the tool itself operates?
How VA Horizon Avoids This Risk Category Entirely
Human + AI SDRs run real, individually operated SMS conversations with each prospect, built natively on the VA Horizon Private CRM, not automated LinkedIn outreach of any kind. Every conversation is a real, trained VA texting a real prospect, which means the specific ToS exposure this guide describes simply does not apply to how your meetings get booked. That is a structural difference, not a policy promise: the risk this guide covers requires a tool operating on LinkedIn's platform, and VA Horizon's delivery channel is not that.
What this means for you
- LinkedIn's User Agreement bans automated outreach in plain, verbatim, broadly written language covering scraping, bot-driven connection requests, and automated messaging alike.
- Trade press reported a real 2025 enforcement example tied to an AI-SDR company's own team accounts, a useful, honestly attributed data point on how real this risk actually is.
- A vendor whose top-of-funnel depends on LinkedIn automation carries a platform-level risk that a manually operated, human-run channel does not.
Sources
The external data in this guide draws on the sources below. Figures described in the text as estimates or industry triangulations are directional and are not attributed to a single dataset.
- LinkedIn User Agreement
- Artisan AI: from "Stop Hiring Humans" to LinkedIn bans (Complete AI Training)
- Artisan AI: the most controversial AI SDR in tech (Quasa)