AI-generated video for B2B firms: film the trust, generate the information

A B2B firm can use AI-generated video on one condition, and it settles almost every case: the shot may only carry information, never build trust. A diagram of a part moving through a production cell, an animated title, a map of the regions you serve: those can be generated. Your plant, your employees, your machine running, your client speaking and your leader: those get filmed. An industrial buyer has walked real shop floors. He can tell the difference. And once he sees it, he does not only doubt the generated shot. He doubts every other one.

This page is for the owner or marketing lead of a Québec industrial SME who has received a video quote where some shots will be generated, or who wonders whether it is worth trying. It does not tell you which tool to use. It gives you the question to ask of every shot, and what it costs when nobody asks it.

The question that settles it: trust or information

Every shot in a corporate video does one of two jobs.

A trust shot is evidence. It tells the buyer: this exists, this happened, these people are real. The assembly line running, the welder at his station, the part coming off the press, a client's operations director describing the project. The buyer does not watch these shots to learn something. He watches them to check.

An information shot is explanation. It helps someone understand what nobody is asking you to prove on camera. How a heat exchanger works, the order of steps in a process, the map of your delivery points, a spec sheet set in motion. Nobody believes that diagram was filmed. Nobody is misled if it was not.

The test fits in one sentence. If the buyer learned this shot was never filmed, would his opinion of your company change? If yes, it is a trust shot, and it gets filmed. If no, it is an information shot, and how it gets made becomes a question of budget and quality, not credibility.

The trap sits between the two. A mood shot of a generic plant, anonymous hands on a control panel, a spotless warehouse in a wide shot. People think of them as decoration. They are not, because they look like evidence. Placed in your video, they say "this is our place". So they belong in the first category.

What an industrial buyer notices

Take, as an illustration, a prime contractor in Montérégie assessing a machining subcontractor in Beauce. The buyer has walked through dozens of shops. He knows what a machining cell in production looks like: chips, coolant, lockout tags, safety guards, floor markings worn down where the forklifts pass.

A generated plant shot often looks too clean, too symmetrical, with no signage, and with machines that match no model he knows. He does not need to know how the image was made. It is enough to sense that it does not show a real floor. And if he comes to visit, he compares. The video showed him a plant he cannot find.

The real cost is not the shot. It is everything around it. Once a buyer doubts one image, he rereads the rest with the same suspicion. Your client's testimonial, filmed at their site, with their name on it, loses part of its weight because it is cut next to a fabricated shot. That is exactly the asset a B2B video cannot afford to weaken, as I explain in the page on the B2B client testimonial video. A long sales cycle rests on an accumulation of small proofs. One false note is enough for all of them to be reread.

There is also the question of rules. The Competition Bureau states that under the Competition Act, it is against the law to advertise or market something in a way that is false or misleading. In its advice to advertisers, it says not to use photos or illustrations that are different from the product being sold. It notes that courts consider the general impression a representation conveys, in addition to its literal meaning, and that no one actually needs to be deceived for a practice to be found misleading. The pages consulted do not mention AI. The reasoning that follows is ours: a generated plant presented as yours conveys a false general impression of your capacity. For your specific case, that is a question for your own legal counsel.

Where generation belongs

It would be dishonest to claim AI has no place in a corporate video. Several uses never touch trust.

• Illustrating what no camera can film. The inside of a sealed chamber, flow in a pipe, a reaction at molecular scale. An animated diagram, drawn or generated, is still a diagram. Provided it looks like one.
• Preparing the shoot. A first outline, reference images to show the client the framing you are after before the day. None of it is published.
• Graphic elements. Titles, background plates for graphics, the icons in an infographic.
• Subtitles, as a first draft, reviewed by someone who knows your technical vocabulary.

The platforms draw a similar line. YouTube requires creators to disclose realistic content that makes a real person appear to say or do something they did not, alters footage of a real event or place, or generates a realistic scene that did not actually occur. The same page exempts caption creation and production assistance, such as an outline, a script, a thumbnail, a title or an infographic. Creators who consistently choose not to disclose may face a label applied manually, or penalties up to removal of the content.

Read that list with your own project in mind. The realistic scene of a plant that does not exist is exactly what you will be offered to "complete" a shoot. It is also what YouTube asks to be disclosed. A corporate video in which a shot of your floor has to be declared synthetic has already lost the argument it set out to make.

When this rule does not apply

Three situations where the question plays out differently.

Internal training video. A module explaining a lockout procedure to employees who already work in the plant has nothing to prove. They know the floor better than any camera. Generated diagrams can be entirely adequate there, as long as the procedure shown is accurate.

The plant that does not exist yet. A Saguenay manufacturer announcing an expansion can show a rendering of the future building. That is honest, provided it is presented as a rendering. The shot informs about a project, it claims to prove nothing. The day the building stands, it gets filmed.

Clearly stylised content. An animation nobody could mistake for reality passes itself off as nothing. The risk begins the moment the image turns realistic.

What to demand from your producer

You do not need to become an expert in the tools. You need to know what in your video was actually filmed.

• A shot list stating, for each shot, whether it was filmed at your site, taken from a stock library or generated.
• A commitment that no generated shot depicts your plant, your employees, your product, your clients or your leader.
• For any realistic generated shot that remains, a written reason and how it will be identified.
• A clear answer on the rights to generated elements.

That last point is not a formality. According to the federal government's consultation paper on copyright in the age of generative AI, case law attributes authorship to a natural person who exercises skill and judgment. The paper considers it far less probable that this criterion is met for works produced by a generative AI system based solely on short instructions. It adds that who could be the author and first owner of such works remains an open question. In other words, a generated shot may be an asset nobody clearly owns. A shot filmed in your plant by a person does not raise that question.

One answer should end the conversation: "nobody will see the difference". Your buyer is not nobody. He is someone whose job is to spot what does not hold up.

Our position is printed on the home page: certified human content. I shoot the videos myself, on the plant floor, and the person who shoots is the person who then runs the ad accounts where the video will run. We are two people, and nothing passes through a middleman who might "complete" a missing shot without telling you. How we produce is described on the video and content production page, and what a day on the floor requires is in the guide to shooting video in a plant.

Sources

• YouTube, Disclosing use of GenAI content, support.google.com/youtube/answer/14328491, fetched 30 September 2026.
• Competition Bureau Canada, Deceptive marketing practices, competition-bureau.canada.ca/en/deceptive-marketing-practices, fetched 30 September 2026.
• Competition Bureau Canada, General impression test, competition-bureau.canada.ca/en/deceptive-marketing-practices/general-impression-test, fetched 30 September 2026.
• Competition Bureau Canada, Advertising dos and don'ts, competition-bureau.canada.ca/en/deceptive-marketing-practices/types-deceptive-marketing-practices/advertising-dos-and-donts, fetched 30 September 2026.
• Innovation, Science and Economic Development Canada, consultation paper, Copyright in the Age of Generative Artificial Intelligence, ised-isde.canada.ca/site/strategic-policy-sector/en/marketplace-framework-policy/consultation-paper-consultation-copyright-age-generative-artificial-intelligence, fetched 30 September 2026.