AI and authenticity
Should founders use AI avatars or clones for LinkedIn video?
Short answer
We'd advise against it for your own posts. Video earns trust because it is really you, and an avatar gives that away. Audiences are wary of synthetic presenters, and LinkedIn can show content credentials on AI-made video. If recording is the hard part, make recording easier rather than replacing yourself.
19%
of people across six countries say they are comfortable with news made using an artificial presenter or author, in the Reuters Institute's 2025 report on generative AI
What is a founder video actually for?
Showing people who you are. Text can be drafted by anyone, and readers know it. A video of you talking shows how you think on your feet, how you sound and whether you seem like someone they would like to work with.
That is why video builds familiarity faster than text, and why a buyer who has watched a few of your clips can feel they already know you when you first speak.
An avatar removes exactly that. It shows a generated copy of your face saying words you never said, so the thing the viewer is judging is no longer you.
What does the research say about trust?
People are wary of synthetic presenters. In the Reuters Institute's 2025 report on generative AI, surveying six countries, only 19 percent said they were comfortable with news made using an artificial presenter or author. That was about news rather than LinkedIn, but the same instinct applies to a founder's face.
An avatar also leaves you with an awkward choice. You can label it and invite the question of why you didn't record yourself, or leave it unlabelled and hope nobody notices.
Will LinkedIn show that a video is AI-made?
Sometimes. LinkedIn's help page on content credentials says members can see provenance details on images and videos that carry C2PA information, a tag that some cameras and AI tools embed to record how media was made, and that the feature is still rolling out.
It also says LinkedIn can't yet identify and label all AI-generated content, so a missing label proves nothing either way. Plan as if viewers may find out from someone or something other than you.
What if being on camera is the problem?
Then work on that instead of replacing yourself. In our experience, most founders who avoid video don't lack the ability. They lack time, a routine and a way to start without a blank page. A few things help.
- Answer questions rather than presenting, so you are talking the way you do on a sales call.
- Record short clips, one point each, so a stumble costs a few seconds rather than a whole take.
- Book a regular slot, so recording becomes routine rather than an event.
- Start with text posts and add one video a week once posting feels normal.
The first few recordings feel awkward for most people, and it passes with practice. Talking for a few minutes, by voice or on camera, is still far less work than writing and checking a script for an avatar.
Follow-up questions
Is it fine to use AI to edit a video I recorded myself?
Yes. Cutting pauses, adding subtitles, fixing the sound or reframing the shot all leave you saying what you actually said. The line is between tidying a real recording and generating one you never made.
What about an AI voice reading my written post?
It has the same problem in a smaller form. Listeners hear a voice that sounds like you saying words you never spoke, and finding out later feels like a trick. If you want audio, record yourself reading it, which takes a couple of minutes.
Are there any good uses for an avatar?
Possibly for things where nobody expects to meet you, such as a product walkthrough clearly labelled as generated, or translating a video you really did record. For posts meant to show who you are, the real recording is the point.
Talk for a few minutes. Post all week.
Hummbird turns your ideas into LinkedIn posts and videos in your voice. Nothing publishes until you approve it.
Start for free7 days free, or 14 with a card