Third party outreach tools were clunky and unreliable, so we replaced them with a custom AI and scheduled tasks system. That left open the question of what to actually say in cold outreach once AI was doing the writing.
I tested a transparent pitch, one that led by disclosing the AI agent, trained on a large volume of my own real emails and messages, instead of hiding the automation the way most outreach does.
The instinct with AI written outreach is usually to hide it, the theory being disclosure kills trust. I bet the opposite. Name the tool honestly, back it with a real detail that makes the personalization claim credible, and it reads as more trustworthy than a polished but generic message pretending to be personally written.
Early results were positive, a couple genuine responses a day across roughly two hundred emails and a similar volume of LinkedIn messages sent daily. We're still weighing high volume outreach against the smaller, higher touch style that worked well before.
Disclosing you used AI can be the credibility play, not the liability, if you can back it with a real detail.
I bet on honesty over the industry's default instinct to hide the machinery. People respond better to the truth than to a polished disguise.