Invisible Hands

Today, an entire submission for a new gene therapy had been reviewed in under four hours

The Invisible Hands

It wasn’t the speed that surprised me—it was how natural it felt.

For years, regulatory submissions had been a slow and painstaking process. Teams of specialists labored over every sentence, every data table, every response to agency questions. The FDA, in turn, would spend months reviewing, debating, and scrutinizing every claim.

But today, an entire submission for a new gene therapy had been reviewed in under four hours.

---

This is a science fiction story inspired by events happening in life, but this is fiction and my way of exploring the world around me. I’d love your thoughts and feedback!

---

I skimmed through the report. The AI had prepared everything: efficacy data cross-referenced with historical trials, safety modeling for long-term risks, and real-time comparisons against thousands of similar treatments. On the agency side, our AI ran its own analysis, benchmarking outcomes, simulating patient populations, and stress-testing the therapy against edge-case scenarios.

It was clean. Thorough. No errors. No missing pieces.

The drug was going to save lives.

And yet, here I was, one of the last human reviewers in the loop, reading through the file before final approval.

I leaned back in my chair, glancing at my colleague across the desk. “Feels too easy, doesn’t it?”

Dr. Patel chuckled. “You remember what this used to be like? Six-month turnarounds. Endless clarification letters. Meetings about meetings.” She tapped her screen. “Now we can focus on the real questions.”

She was right.

The AI had taken care of the mechanics—the formatting, the citations, the regulatory cross-checks. But humans still had a job to do.

Because the role of a regulator was never just about enforcement. It was about judgment.

The System at Work

The AI revolution hadn’t just made things faster, it had made things better.

Patients were getting life-saving treatments in record time. Rare diseases that had once been untreatable now had promising therapies that reached the market before a patient’s condition could worsen beyond repair. Clinical trials adapted in real-time, optimizing for success rather than rigid protocols.

But at the heart of it all, humans still mattered.

On the sponsor side, regulatory teams still set the strategy. AI could write, but it couldn’t decide what should be written. It could structure a perfect response, but it couldn’t determine the best way to frame the science for a skeptical audience. That was still a human decision.

At the agency, AI could detect patterns and flag inconsistencies, but it couldn’t replace expertise. It couldn’t decide when to push back, when to demand more evidence, or when to ask if a treatment’s benefits really outweighed its risks. That was still our job.

Automation had cleared away the bureaucracy, leaving us with the part that mattered most: thinking.

---

A Human in the Loop

I turned back to the submission. The AI had flagged a section for human review, not because there was an error, but because it recognized that this was a gray area.

The therapy used a novel delivery mechanism, something without a direct regulatory precedent. The AI had modeled the risks, but it couldn’t account for nuance, it didn’t know how regulators might feel about a treatment that worked in theory but carried an unknown social or ethical weight.

That was why I was still here.

Because regulatory science wasn’t just science. It was policy. It was trust. It was about making sure that when a patient received a new treatment, they could be confident that someone, someone human, had made the call to approve it.

I added my comments, requested a discussion with the review team, and sent the report back through. Not as a delay, but as a safeguard.

The drug would still be approved in record time.

A child with a rare disorder would still get the treatment before their condition became irreversible.

But it wouldn’t be an AI making that decision alone.

It would be people.

And in the end, that was what the FDA had always been about, not just speed, not just safety, but trust.

Because in medicine, people don’t just need solutions. They need to believe in them.

And that, no matter how advanced the technology, would always be a human responsibility.