The announcement of the Base Editing Access Program — a collaboration between Revvity and Profluent offering researchers access to AI-engineered adenine base editors — represents a significant moment in our industry. For the first time, sophisticated gene editing capabilities previously available only to well-funded institutions are being offered to a broader research community.
We welcome this development. We also have concerns.
The Promise
Let us be clear about what democratization offers: faster innovation, diverse perspectives, solutions to problems that large institutions might overlook. Some of the most significant advances in biotechnology have come from unexpected sources — small labs, academic researchers, even patient advocates who refused to accept that their conditions were untreatable.
Lowering barriers to entry means more minds working on humanity’s most pressing biological challenges. This is unambiguously good.
The Complexity
Gene editing is not software development. The consequences of errors cannot be patched in the next release. A poorly designed edit doesn’t crash a system — it crashes a human being.
The AI-enhanced tools being democratized are powerful precisely because they lower the expertise threshold required to design effective edits. But designing an edit and understanding its implications are different competencies. An AI can optimize for on-target efficiency while remaining agnostic about tissue-specific expression patterns, chromatin accessibility variations, or the downstream effects of altering one node in a deeply interconnected biological network.
These are not theoretical concerns. Our safety databases contain case studies — shared confidentially among institutional review boards — of interventions that appeared optimal in silico but produced unexpected results in vivo. Results that, in some cases, could not be reversed.
Our Perspective
At Unzyme Laboratories, we’ve invested heavily in what we call “enhancement ecology” — understanding each modification not in isolation but as part of a living system. This requires infrastructure that cannot be replicated by access to editing tools alone: longitudinal monitoring capabilities, adverse event databases spanning thousands of subjects, computational models trained on proprietary outcome data.
We do not share this to discourage newcomers to our field. We share it to emphasize that the tools are only part of the picture.
A Proposal
Rather than gatekeeping access to technology — which we neither advocate nor believe is ultimately possible — we propose an alternative: federated safety networks.
Institutions and researchers using these newly accessible tools should be required to contribute outcome data to shared databases. Not proprietary sequences or commercial secrets, but safety signals: what worked, what didn’t, what surprised us. This is standard practice in pharmaceutical development. It should become standard practice in genetic enhancement.
Unzyme Laboratories is prepared to open our adverse event database to qualified researchers who commit to reciprocal sharing. We invite other established organizations to join us.
The Stakes
The next five years will determine whether genetic enhancement becomes a regulated, trusted, equitably distributed technology or a fragmented landscape of variable quality and unpredictable outcomes. The tools being democratized today will be used by thousands of researchers by 2028. What frameworks will guide their work?
Democratization without infrastructure isn’t liberation. It’s abandonment disguised as empowerment.
We can do better. We must do better.
Erik Lindqvist is Chief Executive Officer of Unzyme Laboratories.
Unzyme Laboratories’ Safety Data Sharing Initiative launches Q3 2026. Interested institutions may register interest through our Contact page.