In this blog, Natalia Kaczmarek, Whiteoaks’ Digital Content Manager, discusses:
- The growing concerns about the volume of AI-generated content on LinkedIn
- What LinkedIn’s new feedback feature ‘Seems like AI slop’ is about
- How B2B brands can protect the quality of their LinkedIn content
The chatter about AI copywriting on LinkedIn has been going on for quite a while, but what happened on my feed during the final week of July could only be compared to a digital feeding frenzy. Everyone had an opinion and most had the same screenshot.
So what’s this all about?
Well, LinkedIn has started rolling out a new feedback feature with rather blunt wording: ‘Seems like AI slop’ – and it went, dare I say… viral?
Hidden within the three-dot menu on posts, the button allows users to flag content they believe is overly reliant on AI. Given the amount of suspiciously polished ‘thought leadership’ filling the platform, it is easy to see why many fellow LinkedIn users have welcomed it.
After all, LinkedIn wants its feed to prioritise genuine expertise, lived professional experiences and original viewpoints on industry challenges. Let’s be honest, we’ve all been craving that authenticity and uniqueness as LinkedIn over the last year or so has not always delivered that.
Giving people more control over what they see could reduce recycled opinions and automated comments that are saying absolutely nothing. So, on the surface, the new ‘Seems like AI slop’ feature sounds sensible and like the right step forward for the platform.
But there is a ‘but’, or rather a few ‘buts’…
Let’s start with the feature’s name
Call me crazy (reserved for colleagues only :-)) but the name of the feature itself seems totally off-brand for LinkedIn. For a rather corporate-sounding, professionally-minded brand, the name has definitely raised some eyebrows – mine too – because it feels so informal.
Although, according to the Cambridge Dictionary, once you scroll past the original definition involving wet food waste, ‘slop’ can also mean “content on the internet that is of very low quality, especially when it is created by artificial intelligence”. It does come with the caveat that this use is informal.
So, LinkedIn has used the word correctly. It just sounds more like something you would see in the group chat than expect to find within a LinkedIn product menu.
Then, there’s the timing
There is some comic timing involved too because LinkedIn, not so long ago, introduced a tool that could help you rewrite your posts using AI. It is now replacing that feature while asking users to report copy that feels AI-generated.
That has understandably confused parts of the social media community. It appears the platform first handed everyone the robot pen and has now introduced a button for complaining about the handwriting.
Of course, using AI does not automatically result in AI slop. The problem comes when the tool replaces original thinking rather than supporting it.
AI slop is subjective
The bigger issue, though, is that ‘AI slop’ is subjective. A LinkedIn post may have been copied directly from someone’s favourite AI tool, carefully written by a human or created by someone whose natural style happens to be polished and structured.
From my own experience, anyone who learnt English as a foreign language at school may remember being encouraged to use contrastive negations and other complex structures in essays and exams to demonstrate the depth and breadth of their good old English grammar.
Now, some of those structures are regularly presented as tell-tale signs that somebody has used AI. And don’t get me wrong, repeated contrastive negations can be a giveaway, alongside the somehow still ever-present rocket emoji.
However, without clear criteria, people may end up reporting content they dislike rather than content that is genuinely low quality or full of recognisable AI copywriting habits. Human writing could be mistaken for AI, while better-edited AI content could pass unnoticed.
There is also a danger that the conversation becomes too focused on whether AI was involved rather than whether the content is actually useful. A post can be written entirely by a human and still be repetitive, vague or lacking in substance. Equally, AI may have helped someone organise a genuinely original argument, improve their grammar or communicate more confidently in a language that is not their first.
The presence of AI does not automatically make content slop. Equally, the absence of AI does not automatically make content good.
Could it cause unethical practices?
We also cannot ignore the risk of misuse here. Could competitors repeatedly flag a company’s posts in an attempt to damage their reach? Could leaders using English as not their first language be unfairly penalised for using AI to refine their wording? Could perfectly good content be flagged simply because it uses phrases people have started to associate with AI?
At the moment, we simply do not know.
The information available at the time of writing does not explain whether receiving one or more ‘AI slop’ flags will reduce a post’s distribution. It also does not establish whether reports could affect future posts from the same person or company page.
There is currently no evidence that competitors can use the feature to reduce another brand’s reach. I would hope professional ethics would prevent that behaviour. But until LinkedIn provides more information about how the feedback is assessed and what safeguards are in place, these are fair questions to ask.
What should B2B brands do now?
To start with, copying and pasting the first output that ChatGPT gives you into LinkedIn has always been a poor content strategy. Every post should have human oversight, include an original opinion, relevant experience or useful detail and data that could only have come from the person or organisation publishing it.
That does not mean brands and PR professionals should stop using AI. As we have said before at Whiteoaks, AI tools can help with research, ideas and structure and it can give writers or subject matter experts a useful starting point when the blank page is winning.
The audience-facing copy still needs to be checked, shaped and edited by someone who understands the subject, the audience and the tone of voice.
The strongest LinkedIn content starts with a real experience, such as a customer conversation at an industry event, a lesson from a recent project or a clear view on an industry issue. AI can help frame that thinking, but it cannot supply lived professional experience that was never provided in the first place.
That human oversight and editing is so incredibly important right now as the mystery surrounding the new feature continues.
Whatever LinkedIn decides to do with the feedback it receives, the reaction to the new button tells me one thing – that people are tired of generic content.
Our digital and content teams help B2B tech brands turn real expertise into LinkedIn content that sounds like the people behind it and supports wider business goals. Speak to the Whiteoaks team about bringing more human insight into your feed.
