AI-written LinkedIn posts signed by your CEO: whose words are they

A LinkedIn post signed by a company leader is read as that leader's word, whoever typed it. An AI tool can correct, shorten or translate what the leader said. It cannot supply what the leader thinks. LinkedIn's published community policies do not target writing tools: they target deception, false identity and undisclosed synthetic media. So the real risk is not the platform. It is the buyer who meets you after reading, under your name, a post you would not recognise.

This page deals with one question only: who writes the leader's words. Choosing the account and the pace is covered in SME leader visible on LinkedIn. Choosing a theme is covered in B2B SME thought leadership.

What a buyer reads into a post signed by a CEO

Take, as an illustration, a machined-parts manufacturer in Saint-Hyacinthe trying to get onto a prime contractor's approved supplier list. Before the first meeting, the buyer opens the president's profile. There is one post a week. Lists of five leadership lessons, short lines stacked on top of each other, a question at the end to start the conversation. Nothing about a part, a tolerance, a customer or a machine. Nothing the president of an accounting firm in Laval could not have posted.

The buyer does not wonder whether a tool wrote these posts. He asks something else, and it is worse: does this person have anything to say about my problem. A generic post signed by a CEO answers no, in public, every week.

The second cost comes in the meeting. The buyer mentions a post. The president does not remember it, or defends it badly because he never thought it. The gap between the account and the person becomes visible, and the person pays for it, not the tool.

A buyer does not judge the tool. He judges the person whose name sits at the top of the post.

What a credible post contains, by contrast, comes from a specific place. A contract turned down, and why. A mistake on a job site in the Beauce, and what it changed in the way the firm bids. An open disagreement with a practice in the industry. A tool can shape those sentences. It cannot produce them without inventing them, and an invented experience under a leader's name is a false statement.

What LinkedIn's rules say, and what they do not

LinkedIn's Professional Community Policies set three requirements that bear directly on the question:

• do not share content that is false, misleading or intended to deceive;
• use your true identity and provide accurate information about yourself and your organisation;
• do not share synthetic or manipulated media that shows a person saying or doing something they did not say or do, without clearly disclosing that it is fake or altered.

What is missing matters just as much. The page does not say that posts written with an AI tool are banned or penalised. It targets deception, not the tool. If someone tells you LinkedIn punishes generated posts, ask them for the official page that says so.

LinkedIn also shows a provenance icon on images and videos that are cryptographically signed under the C2PA standard. Clicking it can show whether AI was used to generate or edit part of the content. That mechanism covers images and video, not the text of a post. And LinkedIn itself states that it is not yet possible to identify and label all AI-generated or AI-modified content.

Our reading, which is not legal advice: the tool is not the problem, the content is. A generated post describing a scene the leader never lived, something like "yesterday on our shop floor an operator told me", is a false statement made under his real name. It looks exactly like what the first rule targets, however it was drafted. A post tightened by a tool from what the leader actually said has no such problem.

Where a tool helps, and where it costs trust

The line I draw fits in one sentence: the substance comes from the leader, the form can be assisted.

A tool belongs where it works on material that already exists:

• transcribing a voice memo recorded in the truck between two plants;
• tightening a draft that runs long, or fixing spelling;
• preparing an English version for American buyers of a post the leader wrote in French, then having it read by someone who knows the trade;
• suggesting a few headlines for the leader to choose from.

It costs trust when it supplies the material itself: an opinion the leader never voiced, an anecdote, a figure, a customer, or an imitation of his voice on a subject he has never spoken about. It also costs trust when the post goes out before the leader has read it.

The test is simple. Could the leader say the post again, in his own words, to the buyer who quotes it back to him? If yes, how it was shaped hardly matters. If no, it should not carry his name.

What to demand from anyone who writes for the leader, whether an assistant, a copywriter or an agency:

• every post starts from something the leader said, and you can say where and when;
• the leader reads everything before it goes live and can cut without argument;
• no anecdote, no figure and no named customer that did not come from him.

The answer that should end the conversation: "we have a tool that has learned your style." A style is not a point of view, and the point of view is what the buyer came for.

When the question does not really apply

There are cases where this page does not apply, and I would rather say so.

Factual notices posted from the company page: a new certification, a job opening at the Saguenay plant, a new machining centre on the floor. Nobody reads a personal conviction into a notice. A tool drafting it deceives no one, as long as the facts are right.

The leader who dictates and has it corrected. A president who works in French and posts in English for an American market can have the text translated and polished. The idea is his, the facts are his. That is help with form, not substitution.

The firm whose buyers do not look at LinkedIn. A supplier that sells almost entirely through public tenders, where price and compliance decide, has other priorities than who holds its president's pen.

And the leader who does not post at all. The question then is not who writes, but whether to start, and that is not the subject of this page.

How we approach a leader's words

The badge on our home page reads "Certified Human Content", with the line "NOT PROMPTS, NOT SLOP. CRAFT." That does not mean a tool never has a place. It means a leader's word is not something you generate.

In the Signal Program, the leader spends half a day every 3 months in front of the camera. What he says during that half day becomes the material: three videos of about a minute and six clips of 15 to 30 seconds, in every format, distributed to the CEOs, VPs and directors of his market. In those videos every word is his, with his face and his voice. And because we are two people, and the person who shoots is the same one who runs the ad accounts, nothing he says passes through a middleman who would rewrite it.

If your real question is frequency or which account to post from, go back to SME leader visible on LinkedIn. If it is the subject to hold over time, it is in B2B SME thought leadership.

Sources

• LinkedIn, Professional Community Policies, https://www.linkedin.com/legal/professional-community-policies, fetched 30 September 2026.
• LinkedIn Help, Content credentials, https://www.linkedin.com/help/linkedin/answer/a6282984, fetched 30 September 2026.