// AI NEWS · PLATFORM

LinkedIn Reports a 46% Jump in Detected Inauthentic Activity as It Tightens Enforcement

LinkedIn's latest EU DSA disclosure shows detected inauthentic activity up 46% in the first half of 2026, with engagement pods, third-party auto-posting services, and AI-generated engagement all in scope.

2026-09-02 · by Moe Ameen

What happened

LinkedIn's latest transparency filing under the European Union's Digital Services Act (DSA) reports a 46% increase in detected instances of inauthentic activity during the first half of 2026 compared with the second half of 2025. The disclosure, surfaced in early September 2026, frames the jump as a result of stepped-up detection rather than a claim that the platform is being overrun — LinkedIn is finding and acting on more of it.

The activity in scope is broad. LinkedIn names automated engagement, artificial-intelligence-generated content ("AI slop"), and suspicious or inauthentic profiles, and it specifically calls out engagement pods — groups that trade coordinated likes and comments to game the feed — and third-party services that automate posting and interactions. This is the same enforcement surface behind LinkedIn's member-facing "AI slop" report button and its wider push on automated content over 2026.

The same report puts LinkedIn at 56.5 million active users in the EU for the first half of 2026, up about 1.4 million from the prior period. That context matters: the enforcement increase is not being driven by a collapse in real usage, and roughly a third of the EU member base uses the app in a given period — consistent with earlier filings.

One distinction is worth keeping straight, because it decides who is actually at risk. LinkedIn's enforcement is behavioral, not a blanket ban on automation. It targets tools and accounts that mimic bot-like behavior — fake engagement, session automation, mass connection or DM blasts, and low-effort AI content pushed at volume — not the ordinary act of scheduling original posts. Treat the exact figures and category names as reported in the DSA disclosure; confirm current numbers against LinkedIn's own transparency pages before relying on them.

Why it matters for creators

  • If your growth leaned on engagement pods or a tool that likes, comments, or sends connection requests on your behalf, that is exactly the behavior the 46% increase is catching. The risk of a reach penalty or account action is rising, not falling.
  • Third-party services that automate posting by driving your logged-in session (browser bots) are a different risk class than API-based publishing. The former looks bot-like to LinkedIn's detectors; the latter is how scheduling is supposed to work.
  • AI-generated content is in scope when it reads as slop — high volume, low effort, generic. The lesson from LinkedIn's enforcement year is that using AI is fine; publishing obvious, undifferentiated AI output at scale is what gets flagged.
  • Enforcement is behavioral, so the safe path is not "post less" — it is "post original, on-brand content through sanctioned channels and stop faking engagement." Consistency of real posting is not what LinkedIn is hunting.
  • This is a durable signal, not a one-week story: DSA reporting recurs, and the trend line across 2026 is more detection, faster flagging, and less tolerance for artificial engagement across LinkedIn and its peer platforms.

How to act on this with Kompozy

The trap creators fall into after news like this is to conflate two very different things: automating engagement, and automating publishing. LinkedIn is cracking down on the first — pods, fake likes, mass DM and connection bots, and tools that puppet your logged-in session to look human. It is not banning the act of drafting a post and scheduling it. Kompozy sits cleanly on the safe side of that line, and that is the whole point of using it instead of a grey-hat automation stack.

[Kompozy](/) does not touch engagement. It never likes, comments, follows, or DMs on your behalf, and it does not drive your browser session. What it does is generate original content — talking-head [Persona Shorts](/glossary/persona-shorts), Carousels, Photo Posts, Text Posts, Blog Articles, and newsletters, each held to one [Persona Brief](/glossary/persona-brief) so the output reads as a real voice rather than volume slop — and then publish it through sanctioned integrations rather than session bots. LinkedIn posting runs through official partner infrastructure, and [Autopilot](/glossary/autopilot) schedules the set across the eight social platforms plus blog and email behind a per-post review you control. The reader who just learned their pod or auto-liker is now a liability gets the honest answer here: keep the real thing — original content, published on a real cadence — and drop the behavior LinkedIn is actually flagging.

Quick takeaways

  • LinkedIn's EU DSA disclosure, reported in early September 2026, shows detected inauthentic activity up 46% in H1 2026 versus H2 2025.
  • In scope: automated engagement, engagement pods, third-party auto-posting/interaction services, AI-generated "slop," and suspicious profiles.
  • The platform reported 56.5 million active EU users for H1 2026, up ~1.4 million — so the crackdown is not being driven by falling real usage.
  • Enforcement is behavioral: it targets bot-like engagement and session automation, not the ordinary scheduling of original posts.
  • The durable move is to stop faking engagement and publish original, on-brand content through sanctioned channels — which is exactly how Kompozy works.

Frequently asked questions

What did LinkedIn report about inauthentic activity in 2026?

In its EU Digital Services Act disclosure surfaced in early September 2026, LinkedIn reported a 46% increase in detected instances of inauthentic activity in the first half of 2026 compared with the second half of 2025. The categories include automated engagement, engagement pods, third-party auto-posting services, AI-generated "slop," and suspicious profiles.

Does this mean LinkedIn is banning all automation and scheduling?

No. LinkedIn's enforcement is behavioral — it targets tools and accounts that mimic bot behavior, such as fake likes and comments, engagement pods, mass connection or DM blasts, and services that automate your logged-in session. Scheduling original posts through sanctioned publishing is a different, permitted activity.

Is publishing AI-generated content on LinkedIn now risky?

Using AI to help create content is not itself the problem. The risk is publishing obvious, generic AI output at high volume — the "AI slop" LinkedIn and its members flag. Original, differentiated content that happens to be AI-assisted, posted at a normal cadence, is not what the enforcement targets.

How can I keep automating my LinkedIn workflow safely?

Separate engagement automation (risky) from content automation (fine). Do not use tools that like, comment, follow, or DM for you or that puppet your browser session. Do use a generation engine that creates original posts and publishes them through official integrations on a real schedule — that is the model Kompozy follows.

Related news

← All AI news · Get started →