How generative AI has changed LinkedIn publishing and what it means for creativity

 

How generative AI has changed LinkedIn publishing and what it means for creativity by D. Conterno

Executive summary

  • Generative AI has materially altered LinkedIn publishing workflows by reducing the cost (time, effort, confidence) of drafting, editing, translating and formatting posts, especially for “thought leadership” style content. LinkedIn itself now offers an AI-powered writing tool for creating posts and explicitly nudges users to edit and add their own thoughts before publishing.  https://www.linkedin.com/help/linkedin/answer/a1517763
  • Hard, platform-wide year-by-year statistics do not exist publicly for “posts that are >50% AI-generated” because
    • (a) LinkedIn says it does not track whether posts were written/edited with AI
  • The strongest publicly cited quantitative evidence comes from a third-party detector-based study (Originality.ai) that analysed 8,795 public English-language LinkedIn posts ≥100 words from Jan 2018–Oct 2024. It estimates that 54% of long-form posts in Oct 2024 were “likely AI-generated/AI-assisted” (as the detector defines it).  https://originality.ai/blog/ai-content-published-linkedin
  • Evidence from creativity research suggests a trade-off: generative AI can raise individual output quality/creativity ratings (especially for lower-skill writers) but can also reduce diversity/novelty at the collective level by nudging many people toward similar phrasing and ideas.  https://www.science.org/doi/10.1126/sciadv.adn5290

 

1. What has actually changed on LinkedIn since generative AI went mainstream

A. Writing has shifted from “compose” to “orchestrate”
Many professionals now treat LinkedIn posts as something they prompt, shape, and polish, rather than draft from scratch. LinkedIn’s own product design supports this: its AI-powered writing tool for posts starts from a user’s bullet points and produces a draft, then reminds the user to review and add their own thoughts.  https://www.linkedin.com/help/linkedin/answer/a1517763
B. Posting is easier to scale (frequency, length, consistency)
A well-cited external analysis (Originality.ai) reports that after ChatGPT’s public launch (late 2022), long-form LinkedIn posts in their sample became more AI-assisted and average long-form post length increased by 107% (meaning ~2.07× the earlier average, as reported by that study).  https://originality.ai/blog/ai-content-published-linkedin
C. The “LinkedIn voice” has become more templated and homogeneous
The same WIRED reporting that surfaced the Originality.ai analysis notes that AI-written corporate-style prose can be difficult to distinguish from human content on LinkedIn, partly because the platform already rewards polished professional language. https://www.wired.com/story/linkedin-ai-generated-influencers/
D. A larger share of users can publish confidently (including non-native English speakers)
WIRED reports that several non-native English speakers use AI tools to polish grammar and translate/rework drafts, lowering barriers to participation.  https://www.wired.com/story/linkedin-ai-generated-influencers/
E. A supporting ecosystem has formed around “AI LinkedIn content”
WIRED also describes a “cottage industry” of tools that generate posts and comments for LinkedIn (for example, dedicated post/comment generators). https://www.wired.com/story/linkedin-ai-generated-influencers/

 

2. Statistics

The important key limitation

  • LinkedIn says it does not track how many posts are written or edited with AI tools. https://www.wired.com/story/linkedin-ai-generated-influencers/
  • Because of that, any numbers you see publicly are estimates based on samples, surveys, or detector outputs, each with material uncertainty.

Best publicly cited quantitative evidence specific to LinkedIn long-form posts

Originality.ai study (published 28 Oct 2025; data through Oct 2024):

  • Dataset: 8,795 public LinkedIn posts (English-language), ≥100 words, spanning Jan 2018–Oct 2024. https://originality.ai/blog/ai-content-published-linkedin
  • Estimate: 54% of long-form posts in Oct 2024 were “likely AI-generated” (per their detector’s definition, which includes AI-generated text even if human-edited). https://originality.ai/blog/ai-content-published-linkedin
  • Change event: a 189% increase in suspected AI usage from Jan to Feb 2023 (this is a relative jump, not a platform-wide count). A 189% increase means: if Jan is indexed at 100, Feb is 289. https://originality.ai/blog/ai-content-published-linkedin
  • Engagement claim: posts classified as “likely AI” received 45% less engagement (likes + comments) than “likely original” posts in their post-Dec-2022 subset. https://originality.ai/blog/ai-content-published-linkedin

Yearly statistics about posts that are >50% AI-generated

Getting yearly values for that metric is impossible because:

  1. LinkedIn does not publish or track it (per LinkedIn’s statement). https://www.wired.com/story/linkedin-ai-generated-influencers/
  2. Even the best-known public study above classifies posts as likely AI vs likely human; it does not measure what fraction of each post is AI-written. https://originality.ai/blog/ai-content-published-linkedin
  3. AI detection is an imperfect science; detectors can produce false positives/negatives and may be biased (for example, against non-native English). https://hai.stanford.edu/news/ai-detectors-biased-against-non-native-english-writers

A cautious “year-by-year” view using only verifiable statements

This is the closest defensible summary we can give from cited sources:

Year

Confirm the share of LinkedIn posts that are “>50% AI-generated”?

What we can say, with sources

2018–2022

No.

In Originality.ai’s long-form sample, AI usage was described as low/negligible before ChatGPT’s release. https://www.wired.com/story/linkedin-ai-generated-influencers/

2023

No.

Originality.ai reports a 189% jump in suspected AI usage from Jan→Feb 2023 in their sample, after ChatGPT’s late-2022 release. https://originality.ai/blog/ai-content-published-linkedin

2024

No (for your “>50% of each post” metric).

For long-form posts ≥100 words, Originality.ai estimates 54% were likely AI-generated in Oct 2024, and reports the “over half” pattern during mid-to-late 2024 in that sample. https://originality.ai/blog/ai-content-published-linkedin

2025

No.

I cannot confirm any credible, platform-wide 2025 estimate for LinkedIn post AI share. LinkedIn still states it does not track AI-authorship of posts. https://www.wired.com/story/linkedin-ai-generated-influencers/

 

3. Is this good or bad for human creativity?

What the evidence suggests (not ideology)

A. AI can raise individual performance and lower the barrier to publishing
Controlled research on professional writing tasks found that access to ChatGPT reduced completion time by ~40% and improved quality by ~18% in one well-cited experiment (not LinkedIn-specific, but directly relevant to writing workflows). https://www.science.org/doi/10.1126/science.adh2586

B. AI can increase perceived creativity but reduce collective novelty/diversity
A peer-reviewed study in Science Advances found that access to generative AI ideas made stories rate higher on creativity/enjoyment, especially for less creative writers, but the AI-assisted stories became more similar to one another. https://www.science.org/doi/10.1126/sciadv.adn5290
A 2025 study likewise reports a pattern where generative AI can enhance creative performance while eroding content diversity over time. https://www.sciencedirect.com/science/article/abs/pii/S0160791X25002775

C. Over-reliance risks “creative atrophy” and shallower thinking in some contexts
There is growing reporting and debate that heavy reliance on AI for writing can reduce cognitive engagement in certain tasks; however, conclusions depend on study design and context. (This is a developing area; treat claims carefully.)

D. Conscious Enterprises Network assessment

Opinion: It is neither “good” nor “bad” in the abstract. It is good when AI is used to amplify human intent (structure, clarity, translation, first draft), and bad when AI is used to replace human intent (generic thought leadership, low substance posting, performative volume).

For LinkedIn specifically, the reputational risk is that AI accelerates a drift toward bland sameness and the data point that “likely AI” long-form posts got lower engagement in the Originality.ai sample is consistent with audiences sensing that sameness. https://originality.ai/blog/ai-content-published-linkedin

 

4. Practical recommendations for LinkedIn publishing (brand + trust + reach)

A.     Use AI, but do not outsource authorship. Use it for outlines, tightening, translation, headline variants… Then ensure the final post contains distinctive substance: an insight, a real example, a number, a source, a stance. (LinkedIn’s own UX nudges this: “add your own thoughts”.) https://www.linkedin.com/help/linkedin/answer/a1517763

  1. Anchor posts in verifiable claims. Where you cite facts, link them to credible sources (institutions, peer-reviewed research, primary documents). This is a competitive advantage in an AI-saturated feed.
  2. Avoid “template tells”. If a paragraph could have been written for anyone, it will read as AI, even when it is human. Keep lived experience, concrete detail, and un-copyable specificity.
  3. Optimise for human resonance, not volume. The marginal cost of posting has dropped; the marginal value of each post has not. If you post less frequently but with higher signal, you differentiate from AI-amplified noise.
  4. Set a simple internal policy: “AI can draft; humans must fact-check, add original perspective, and approve.” This protects credibility, especially for complex content.

Comments

Post a Comment

Popular posts from this blog

The War Against Humanity by D. Conterno (2025)

Image
  The War Against Humanity by D. Conterno (2025) Introduction Humanity faces an unprecedented war that is not fought with guns and bombs but with tools far more insidious: food, education, technology, health, misinformation, disinformation, and fear. This war seeks to control, divide, and weaken populations, stripping individuals of their autonomy and critical thinking. The mechanisms at play span industries, governments, and institutions, forming a complex web of manipulation. We are all being victims of this war by more than just one of its aspects.  By understanding these aspects, we can begin to reclaim our sovereignty and resist the forces seeking to undermine our collective well-being. This article will not dwell on who historically orchestrated and today maintained this war against humanity, behind the scenes or why. Instead, readers are expected to draw their own conclusions and undertake their own investigations. This article will focus on the following categori...
Read more
Image
  Ten Management Lessons CEOs can learn from Mahatma Gandhi By Ashok Bhatia (2015) Mahatma Gandhi , revered the world over as an apostle of peace and non-violence, led India’s freedom movement. His birth anniversary gets celebrated on the 2nd of October. Are his teachings relevant to the world of commerce and business? Can CEOs of today learn a thing or two from his aphorisms? Here are few of his thoughts which business owners, CEOs and managers might find of some interest. The future depends on what you do today. Managements who care for their brands re-engineer their business processes and ensure sustainable operations. They respect the environment and the aspirations of the local communities. They ensure compliance with local laws. Ethics and values are strictly adhered to. A customer is the most important visitor on our premises. He is not dependent on us. We are dependent on him. He is not an interruption in our work. He is the purpose of it. He is not an outsider in our ...
Read more

The Silver Lining in the Corona Virus Cloud

Image
The Silver Lining in the Corona Virus Cloud By Jiten Patel (2020) The current Covid-19 lockdown has provided a great opportunity for us to read more, update our skills and to refresh ourselves with old favourites. One of mine is the 7 Habits of Highly Effective People by the late Stephen Covey. Originally published in 1989, it has been a constant bedside companion of mine. Covey elucidates the concept of ‘Universal Principles’ going beyond what we might call ‘subjective values’.  Here below, I provide my analysis of how Covey’s 7 Habits (or ‘principles’) are great sign-posts for Inclusive Leaders: 1.      Be Proactive Inclusive Leaders are proactive. They make effective use of data to better understand the demographics of their workforce and their customer base/service users. They take appropriate action to close any identified gaps. 2.      Begin with the End in Mind...
Read more