By Chris Russell, Senior Associate Director, Omnichannel Strategy, Inizio Evoke
According to research, AI could have written this blog four times faster than me [1].
There’s no denying that AI helps us create content at a speed that humans (even my talented copywriting colleagues) simply can’t keep up with.
But it has also never been easier – or quicker – to get things wrong. Especially with the growing number of tools, channels and opportunities we now have to reach people. And the very people we’re trying to reach are nervous about AI:
89% of patients say it is important for AI-generated health responses to be validated by a human expert [2], and 71% of physicians using AI cited accuracy and reliability as their top concern [3].
So, how can AI help us create content without sacrificing accuracy, nuance or trust?
AI can be an incredibly useful assistant. It can help you structure an argument, simplify complex language, adapt approved content for different formats and translate a first draft.
But AI is not an authority. It doesn’t own the evidence, understand the full context or carry responsibility for what gets published – even though it may sound full of confidence.
In fact, its confidence is what makes AI dangerous. It can overstate findings, remove uncertainty and make promotional language feel educational.
This leads me to the most important action we all need to take when working with AI:
We must keep a human expert in the loop. A healthcare comms expert should lead the process, apply their judgment and remain responsible for any content produced.
You can use AI to support you, but your expertise is paramount. As is the expertise of healthcare professionals (HCPs). This is why, at Inizio Evoke, we consult HCPs throughout our creative journey.
A cautionary tale: what can happen when AI is left in charge
‘Vegetative electron microscopy’ is a nonsense term that entered scientific literature through digitization and translation errors. Since then, it has spread into AI training data… and reappeared in later models and 20+ scientific papers [4].
AI has preserved and amplified this mistake – at scale – creating ‘digital fossils’ that publishers, researchers and tech companies have struggled to detect or remove [4].
This is just one of many reasons that AI should support human judgment, not replace it.
The safest approach to AI-assisted content creation:
Give it clear boundaries. Use your communications expertise to draft a highly detailed prompt.
Start with approved source material.
Define the audience, tone, platform and required output.
Tell it explicitly not to add claims, data, citations or treatment advice.
Then review every line.
Remember, speed and reach aren’t your final measures of success
All too often in this fast-paced industry, it can feel like attention is the ultimate currency. But attention alone is not success. Trust is.
That’s why, at Inizio Evoke, we combine AI-powered omnichannel strategy, behavioral intelligence and human expertise to create trusted, meaningful engagement for our healthcare clients.
Get in touch here to explore a more effective and human-centered approach to AI-enabled healthcare communications.
References
[1] NP Digital – AI vs. Human Content Creation Time
Research based on a sample of 744 articles comparing time spent on AI-generated versus human-written content.
https://neilpatel.com/marketing-stats/ai-vs-human-content-creation-time/
[2] Wolters Kluwer – Medical AI Patient Trust Survey
https://www.wolterskluwer.com/en/news/wolters-kluwer-medical-ai-patient-trust-survey
[3] Doximity – Physician AI Study, 2026
[4] ScienceReader – Vegetative Electron Microscopy and AI in Science
https://sciencereader.com/vegetative-electron-microscopy-ai-science/