Better Than a Blank Page

A company had a wall coming. A stack of INDs and BLAs to file that ran well past what the regulatory writing team could physically reach in the time it had. The math did not work, and everyone could see it.

The executive team wanted regulatory to trial AI for first drafts. Regulatory said no. Not once, and not briefly. For twelve months.

Twelve months of no

The resistance was not stupidity. It was a writing team that owns some of the highest-stakes documents in the company, being asked to trust a new tool with exactly the work they are most accountable for. Caution there is a feature, not a bug. But caution and a filing wall do not resolve themselves, and for a year they did not.

It took a member of executive leadership to step in and mandate a trial. That is not a small thing to notice: the change did not start with the team’s consent. It started with Permission, used deliberately, because the stall had become more dangerous than the risk everyone was worried about.

Three days in a room

The company signed an expensive paid trial with a vendor. A six-figure trial. And then, alongside it, we ran a head to head.

I sat in a room for three days with my AI Product Partner, ChatGPT, and NotebookLM, and we drafted first-draft sections of an upcoming IND. Not a demo document. Real sections the team would not otherwise have reached for six months or more, buried under the backlog.

Two things came out of that room. First, our drafts, produced with nothing but our standard enterprise agreements and no six-figure vendor, were far superior to what the paid trial produced. Second, and this was the one that mattered, several members of the writing team were dumbfounded. The drafts were not perfect. They were roughly 60 to 75 percent of the way to a usable first draft. But 60 to 75 percent is not the number that matters. The number that matters is that it is not zero.

The aha

A blank page is the most expensive thing in regulatory writing. It is where the backlog lives. Handing a writer a section that is already 60 to 75 percent of the way there does not replace them. It starts them far past the hardest part, and it gives them their time back for the judgment that actually needs a human.

The team saw that, in their own document, in real time. That was the aha, and after twelve months of no it arrived in an afternoon. The momentum after it was enormous, and it did not come from the mandate. The mandate only bought the demonstration a chance to happen. The demonstration is what changed minds, because a regulatory writer watching a real draft of their own work appear is worth more than any number of assurances that the tool is safe.

What it actually was

Be precise about this, because in regulated work the precision is the whole point. No AI-generated text went to the FDA. None of it. The first drafts were used as framing and ideation, a running start on structure and language, and then the team did the hard work themselves: the sourcing, the argument, the precision, the review, the rewriting. The draft was scaffolding, taken down once the building stood.

Every section went through the standard writing process, and the author of record was human from start to finish. What the tool removed was the busywork at the front of the job. What it never touched was the work that reaches a regulator. In regulated documents that line is not a nicety, and it did not move an inch.

What does not transfer

Copy the lesson, not the setup.

The trial only happened because an executive spent real capital to force it after a year of stall. Without that Permission, the wall just gets closer. A demonstration nobody will authorize never converts anyone.

The six-figure vendor was not the villain, but it was the wrong first move. The capability we needed was already inside tools we already paid for, and nobody had tested that baseline before signing the check. That is the whole point of asking whether to self-serve before you buy: sometimes the honest answer is that you already own it. This is not a rule that vendors never add value. It is a rule that you test your own baseline first, because a paid trial is a strong claim to make before you have proven you cannot do it yourself.

And a 60-to-75-percent draft is a starting point, not a submission. Treating it as finished, especially in regulated work, is how you turn a good tool into a real liability. AI-drafted regulatory content has to go through your full human writing, review, and quality process, and what that process must include is a question for your own quality and regulatory function, not for a case study. The win here was escaping the blank page. It was never skipping the work.

Monday morning

Find the workflow that is drowning in backlog, the one where the team is underwater and everyone can see the wall. Before you sign a large vendor trial, run a real head to head with the tools you already own, on a real document, with someone credible in the room. Keep humans as the authors of record and route every draft through your normal process. Then let the demonstration do what twelve months of argument could not.

This is rung two on the ladder, a single workflow re-run a new way. But watch what it starts, because a team that has felt the blank page disappear rarely stops there.

Cheers,
-Titus