Hey everyone, I’ve been testing Copy.ai for some quick email drafts and blog ideas for my CRM projects. Lately, I’ve been running the outputs through a couple of free AI detectors (like Originality.ai and GPTZero) just to see.
The scores are coming back as “highly likely AI-generated” way more often than a few months ago. I thought the whole point was to make content that doesn’t get flagged.
Is this happening to others? If I use this for client outreach or even internal knowledge base stuff, could this become a real problem for us? I’m thinking about email deliverability or even trust if a client checks.
Still learning…
Trying to figure it out.
I've noticed the same trend with my own workflows. The detectors are constantly being updated to catch new patterns, and tools like Copy.ai are also evolving, so it's a bit of a moving target.
Your concern about client trust is valid, but I'd worry less about email deliverability - spam filters are looking for different signals. The bigger risk is if a client or a prospect runs a check on a piece of content you've presented as "your" work and gets a high AI score. That can definitely erode trust.
My approach now is to treat any raw AI output as a draft that needs a human rewrite. For internal knowledge base stuff, it's probably fine, but for outward-facing communication, I always add my own edits, change up the structure, and inject some personal experience. It takes a bit more time, but it avoids that "generic" feel and usually passes the detectors.
Stay connected