The AI label has gone from a policy footnote to an argument about trust. In August, Anthropic began embedding invisible watermarks in what Claude writes, a move tied to new EU AI Act disclosure rules. Platforms are labeling AI-made content too, Content Marketing Institute reports. We asked four leaders at industrial companies whether they would put an “AI-made” label on their own marketing.
Why we asked
So far the argument has mostly been had by software companies. In September, Content Marketing Institute asked six marketing leaders what the watermark changes. All six work at companies that sell software to marketers, and three said plainly that the label is not the problem.
Industrial buyers start from a harder place. Engineers and technical buyers rate generative AI answers 4.7 out of 10 for trustworthiness, according to TREW Marketing and GlobalSpec’s 2026 State of Marketing to Engineers. In these markets, a wrong number on a spec sheet is not a branding problem. So we took the question to the people who sell flanges, industrial dryers, drywall and playground surfaces.
The question
Engineers and technical buyers rate generative AI answers 4.7 out of 10 for trustworthiness (TREW Marketing and GlobalSpec, 2026 State of Marketing to Engineers). Is AI writing or drawing any of what your customers see from you: the website, product images, videos, posts, quotes? Do you tell them? Would you put an “AI-made” label or watermark on your own marketing, and what would it do to a buyer’s trust?
Four answers
Jake Bean
President and Co-Owner, Western Wholesale Supply · LinkedIn
As president of a regional building materials distributor, precision drives everything we do. We never use generative AI to generate material estimates, pricing quotes, or technical submittals.
When contractors bring us blueprints for CertainTeed acoustical ceilings or steel framing, they rely on exact calculations. An unverified estimate can instantly destroy a builder’s margin on the jobsite.
Adding an “AI-made” label to technical materials or pricing would immediately erode customer trust. In commercial distribution, contractors bet their livelihoods on verified operational accuracy, not generative approximations.
Mike Eaton
Director of Marketing, Carrier Process Equipment Group / Heyl Patterson Thermal Processing · LinkedIn
I use AI as a drafting and organizing tool, not as the authority. For Heyl Patterson content like dryer maintenance resources or MultiDisc processor pages, final claims come from engineers, test-lab data, inspections, and documented equipment experience.
We do not use AI to create final product drawings, performance claims, quotes, or images that imply a real installation. If AI materially shaped something customer-facing, I would disclose it.
An “AI-made” label on technical marketing would not kill trust. Hiding it would. But on specs, drawings, or process visuals, it would make buyers question what was verified.
Landon Olson
Operator, Replay Surfacing Inc. · LinkedIn
We use AI to help with early-stage copywriting and brainstorming, but nothing customer-facing leaves without a human — specifically someone who understands rubber surfacing, recycled materials, and what a playground specifier or municipal buyer actually needs to know.
My engineering background taught me that precision matters. A vague surface spec on a splash pad or playground isn’t a content problem — it’s a safety liability. That line keeps AI in the drafting room, not the proposal.
Would I label it? Honestly, yes — but selectively. If AI helped shape a blog post, I wouldn’t lose sleep. If it touched a technical data sheet or product rendering, I’d want that disclosed. Buyers in construction and recycled materials markets are already skeptical; hiding it costs more trust than admitting it.
Lance Vanover
CEO, Lynco Flange & Fitting · LinkedIn
At Lynco, AI can help draft a first pass on general educational content, but it does not create our quotes, certifications, product specs, or technical recommendations. A wrong heat number, flange class, or material callout is not “content”; it is a job-site problem.
We also would not use AI-made product images where a buyer might mistake them for actual inventory, markings, or condition. Traceability matters too much.
Yes, I would label AI-made marketing if it touched what the buyer relies on. Done honestly, it may reduce polish, but it protects trust.
Where they agree, and where they part
None of the four lets AI produce what a buyer acts on. Bean names material estimates, pricing quotes and technical submittals. Eaton names final product drawings, performance claims and quotes. Olson keeps AI “in the drafting room, not the proposal.” Vanover names quotes, certifications, product specs and technical recommendations.
They part on the label. Eaton, Olson and Vanover would disclose AI where it shapes something customers rely on. Bean says an “AI-made” label on technical materials or pricing would erode trust. Eaton sees the same risk on specs, drawings and process visuals, where he says a label would make buyers question what was verified.
Why it matters at the top of the house
The label debate is about how content was made. These four answers are about who stands behind it. None of them lets AI have the last word on a quote, a spec or an estimate, and three would tell a buyer when AI shaped something the buyer relies on. For an industrial company, the label is the visible part. The line behind it is what a buyer checks.
One move to copy
Each of the four drew his line in one sentence. Bean: no generative AI on material estimates, pricing quotes or technical submittals. Eaton: AI drafts and organizes, while engineers, test-lab data and inspections make the final claims. Olson: AI stays in the drafting room, not the proposal. Vanover: AI may draft general educational content, never quotes, certifications, product specs or technical recommendations. A company that has not drawn its own line has four versions here to start from.
Related: Anthropic Began Watermarking Claude’s Output in August. Marketing Leaders Say the Label Is Not the Problem (The Marketing Table, October 14).
How this was made: The Marketing Table put the question to manufacturers and industrial distributors through MentionMatch on October 6, 2026. The answers were submitted on the record that day, by the contributors or their representatives, and run word for word, without greetings or signatures. Headshots were supplied with the answers, except Mike Eaton’s, which is his LinkedIn profile photo; Lance Vanover is shown by initials. Each company name links to the company’s website. Research: TREW Marketing and GlobalSpec, with Elektor, 2026 State of Marketing to Engineers.
Follow The Marketing Table on LinkedIn for one article a day on why marketing matters in industrial companies.

Leave a Reply