One in four LinkedIn posts now AI-generated, raising trust crisis

Pangram, an AI detection company, released a report revealing that one in four social media posts are now AI-generated, with LinkedIn emerging as the platform where this trend is most pronounced. The shift is visible in the proliferation of polished thought leadership posts and career updates authored by generative AI, often published with minimal human revision.

The question driving the article is whether this represents natural workplace evolution or an authenticity crisis. Vivek Bhargava, Co-Founder of Consumr.ai, frames it as the former, drawing an analogy to art history: once photography was invented, the premium for realistic painting disappeared, and contemporary art rose in its place. Similarly, once calculators became universal, the ability to calculate mentally ceased to be a differentiator, and they were permitted in exams. Bhargava argues that authenticity is not the tool but what you do with it—the prompt choice, the curation, the angle. He emphasises that using AI to polish a post after writing it does not diminish originality, though he acknowledges that asking AI to generate the entire point of view and then publishing it under one's own name is different.

Rishi Sen, Founder & Chief Brand Architect at The Sixth Sen, locates the problem not in AI but in the shortcut mentality. He observes that when Canva, stock footage, and design templates made content production easier, everyone accessed the same starting points. The issue arose when creators treated the template as the finished product. AI operates identically: it can structure an argument, sharpen language, or accelerate the process, but it becomes problematic when the starting point becomes the entire thought. Sen notes that many LinkedIn users are not using AI to enhance genuine beliefs or expression but asking AI to generate the viewpoint itself, making cosmetic changes and publishing under their own name. At that point, the content is polished but not authored—it represents the tool's most probable response, not the individual's lived experience or independent thinking. Sen concludes that this is both natural evolution and the beginning of an authenticity crisis; the technology is not the problem, but the easy way out is. When everyone uses the same shortcut, content becomes familiar, interchangeable, and incapable of creating value.

Vishal Rupani, Co-founder of Sprect.com, compares LinkedIn to Mumbai's street-food culture. Every vada pav stall uses the same ingredients, yet regulars walk past identical stalls to reach their favourite because that vendor deliberately burns the bread slightly—a small, intentional imperfection that becomes the brand. LinkedIn today, Rupani argues, feels like hundreds of stalls that have stopped bothering with that distinguishing touch. He frames the growing reliance on AI not as evolution but crisis, claiming that writing one's own thoughts in one's own words used to signal effort, but that expectation has quietly disappeared, rebranded as productivity for comfort. He points to LinkedIn's renaming of its "Write with AI" tool to "Enhance post" as evidence that the capability remains unchanged despite criticism. Rupani recommends that creators test their authenticity by asking whether a post would stand up if someone privately asked them what actually happened in detail. If yes, AI assisted; if no, AI replaced.

Dr. Kushal Sanghvi, Director of Komerz UK, has observed the platform over 15 years and notes that a significant share of posts now appear AI-assisted and are easy to spot because they no longer sound like the individual behind the profile. Regular readers become familiar with a creator's tone, style, and way of expressing ideas; when those shift, AI-generated posts feel noticeably different. Sanghvi warns that the larger concern is dilution of a creator's unique voice—audiences connect with distinct ways of communicating, and when AI writes everything, readers sense they are hearing an AI layer rather than the person. He emphasises that people connect with those who understand them, communicate naturally, and sound like real people, not machines. If a brand relies entirely on AI-written posts, audiences quickly notice and engage less. Originality, he says, is one of the biggest drivers of engagement; people follow brands and creators because they bring fresh perspectives, personal experiences, and unique insights—qualities AI cannot genuinely produce. He notes that breaking news, company updates, and firsthand experiences perform particularly well on LinkedIn, as audiences naturally connect more with something that happened within an organization or an experience the creator has personally had than with generic AI-generated content.

Rupani draws on nearly 15 years in ad-tech and distinguishing genuine activity from bots. He argues that what exposes bots is not usually a major flaw but a small detail nobody bothered to fake because it was not worth the effort. Authenticity often lies in imperfections—friction, pauses, unexpected actions, or small inconsistencies that serve no obvious purpose. Posts written by genuine humans often include oddly specific details, corrected mistakes, or observations that do not necessarily strengthen an argument but make it feel real. AI-generated writing smooths away those details because it has no reason to include them. Rupani cites Radhika Gupta, Managing Director and CEO of Edelweiss Mutual Fund, who openly spoke about a gold and silver fund launched in 2022 that attracted only ₹12 crore at launch and described it as a "silly product" at the time. That kind of specific, verifiable, and slightly unflattering detail is difficult to fake and easy to fact-check; AI will not produce such stories because it has no personal reputation or career attached. Rupani's rule of thumb is to include one honest, slightly unflattering detail for every three pieces of advice—advice is easy and AI already produces it well, but the personal detail is costly and what people remember.

Rishi Sen concludes that as AI makes competent content available to everyone, competence alone will stop being a differentiator. What will distinguish trusted brands, executives, and creators is the quality of raw material brought to the tool: lived experience, proprietary knowledge, original observations, first-party data, and a point of view shaped by actual decisions. AI can organise these thoughts, sharpen language, or improve presentation, but it cannot independently recreate the context, judgment, and consequences behind a genuinely earned opinion. Human connection will come not from deliberately making content less polished or adding anecdotes to appear authentic, but from specificity—talking about what actually happened, what was learned, what changed their mind, and what they believe others may disagree with. The more content AI produces, the more audiences will value content that could only have come from that particular person or brand. In an AI-first world, trust will not belong to the best-written voice but to the most recognisable and well-earned one.

Vivek Bhargava uses the camera analogy to support his point: when one generates content with AI, every output is different depending on previous prompts, feedback, and creative freedom given. Just because everyone uses a superior camera does not mean everyone becomes a great photographer; everyone may become a better photographer, but when everyone becomes extraordinary, extraordinary becomes ordinary and only the truly exceptional becomes the new extraordinary. With a camera phone today, one can create a photo better than the best photographer of the 1990s, but against the best photographer of 2026, there would be no match. Technology can take you only so far.

The article concludes by raising the question of platform responsibility: should platforms actively label or downrank AI-generated posts, or is the responsibility for transparency ultimately with users and organisations that publish the content? Dr. Shangvi suggests that transparency around AI-generated content will become increasingly important as its use spreads, though the article does not provide a definitive answer.

The report's finding that one in four social media posts are AI-generated reflects a broader tension in professional communication: the tool has become competent enough to produce polished, passable content instantly, but this universal access has inverted the value hierarchy. Where productivity once meant writing faster or clearer, the article suggests it now risks meaning the absence of effort—a rebranding that several experts in the piece call uncomfortable but real.

The voices quoted—from Vishal Rupani's vada pav vendor analogy to Rishi Sen's template-as-finished-product comparison—point to a recurring pattern: technology democratizes a skill (photography, design, now writing), initial abundance looks like a crisis, but the surviving differentiator shifts from the tool to the raw material fed into it. Authenticity, in this framing, is not the absence of AI but the specificity, lived experience, and judgment that only a particular person can contribute. Dr. Kushal Shangvi's observation that audiences are already skilled at detecting style shifts after years of following a creator underscores that this is not a future problem—readers are flagging it now.

LinkedIn's own decision to rename its writing tool rather than remove it suggests institutional awareness of the tension without clear resolution. The article frames the choice as one between transparency (labeling AI posts) and user responsibility, leaving that question open, but the expert consensus leans toward a world where trust accrues to those who visibly choose not to outsource their voice.