LinkedIn has become one of the clearest places to watch AI writing move from novelty to routine. The platform is still used for jobs, professional networking, and keeping up with former coworkers, but its feed now also reflects a broader shift across the internet: automated prose is becoming part of everyday communication.
A new analysis shared with WIRED by the AI detection startup Originality AI found that over 54 percent of longer English-language posts on LinkedIn are likely AI-generated. That does not mean every post is fake or careless. It does mean the boundary between human professional voice and machine-assisted writing is getting harder to see.
What the analysis found
Originality AI scanned a sample of 8,795 public LinkedIn posts over 100 words long. The posts were published from January 2018 to October 2024, giving the company a view of how AI writing appeared to change over time on the platform.
For the first few years in that sample, AI writing tools appeared to have little presence. The major change came at the beginning of 2023, when Originality AI found that likely AI-generated posts had spiked 189 percent. According to Originality CEO Jon Gillham, “The uptick happened when ChatGPT came out.”
Since then, the rise has leveled off. But the result still points to a major change in the way professional social media content is made. A feed that once depended mainly on self-written updates, career reflections, and networking posts now includes a large amount of text that may have been drafted, rewritten, polished, or structured by AI.
Why LinkedIn is a natural fit for AI writing
LinkedIn’s tone has always been different from more casual social media platforms. It is a place where users often try to sound polished, careful, optimistic, and broadly acceptable to employers, clients, colleagues, and recruiters. That makes it easier for AI-generated corporate language to blend in.
The platform itself has embraced AI. LinkedIn Premium subscribers can access in-house AI writing tools that can rewrite posts, profiles, and direct messages. These tools fit neatly into the kinds of writing many users already do on the platform: announcing a role change, commenting on a colleague’s promotion, summarizing a lesson from work, or trying to reach a professional audience.
There is also a wider market around LinkedIn content. The source article describes a cottage industry of AI LinkedIn comment and post generators designed for people who want to grow audiences, impress potential bosses, or reach prospective customers. In that environment, AI does not have to create a new behavior. It can simply make an existing behavior faster.
- Users can draft posts more quickly.
- They can adjust tone before publishing.
- They can polish grammar or structure.
- They can produce routine professional comments with less effort.
The result is a platform where AI writing may feel less disruptive than it would in places where readers expect stronger personal voice, humor, or emotional candor.
How users describe the tradeoff
LinkedIn users who spoke to WIRED described several reasons for using AI tools. Some said they rely more on general-purpose large language models than on LinkedIn-specific products. Content writer Adetayo Sogbesan said she uses Anthropic’s Claude to create rough drafts of posts for clients in the tech industry, then edits them afterward.
For some users, the value is not simply speed. Several non-native English speakers told WIRED they use AI tools to polish English writing and fix grammar errors. Journalist and marketer Çiğdem Öztabak said she has experimented with AI to rework posts originally written in Turkish, her first language, and prefers Claude to ChatGPT.
Those examples show why the debate is not simple. AI writing can make a post feel less personal, but it can also help people communicate more clearly in a professional setting. It can flatten voice, but it can also reduce friction for users who are trying to write in English, manage tone, or get past a blank page.
LinkedIn’s position
LinkedIn says it does not track how many posts are written or edited with AI tools. Adam Walkiewicz, LinkedIn’s head of “feed relevance,” told WIRED that the company has defenses to identify low-quality and duplicate content, including exact or near-exact duplicates, and that it takes action when such content is detected so it is not broadly promoted.
LinkedIn’s position is that AI can support the writing process, but that member ideas still matter most. That distinction is important because AI use can range from light editing to full drafting. A post can be machine-assisted without being empty, but a polished structure does not guarantee originality or substance.
This is the core challenge for LinkedIn’s feed. The platform does not just need to decide whether AI writing exists. It needs to decide which posts deserve attention when more people can generate clean, confident, professional-sounding text at scale.
The backlash is about quality and trust
AI writing remains controversial. Some writers and artists oppose large language models because they are trained on human-created books, websites, and other works without permission or compensation. In that view, AI tools can devalue human writing and reduce professional opportunities for people who make a living from creative work. The source also notes that high-profile lawsuits allege that training AI on people’s art and writing without their knowledge is tantamount to theft.
On LinkedIn, the reaction is mixed. Entrepreneur Zack Fosdyck said some people responded positively to the clarity and structure of AI-assisted posts, while others focused on the fact that AI was involved. Rakan Brahedni, a LinkedIn blogger and founder of a technology advisory firm, said there is clear dislike for content that seems blatantly AI-written, but he also uses a disclaimer when he relies on AI writing tools.
That suggests the real dividing line may not be AI versus human. It may be whether the post feels useful, thoughtful, and honest. LinkedIn has long rewarded polished professional language. AI can produce that surface quickly. The harder part is preserving the original judgment, experience, and point of view that make a post worth reading in the first place.