Bioethics
How we think about the intersection of science, artificial intelligence and the people we serve.
Not just whether something can be done.
Bioethics also asks whether something should be done, and under what conditions. It was born in medicine to answer an uncomfortable truth: technical progress, on its own, does not guarantee that people are treated well.
From there came four principles that still underpin it — autonomy, beneficence, non-maleficence and justice — meant to guide decisions where what is at stake is not a product, but human biology and the trust of the person who shares it.
Neural Omega is not a hospital, but it works with the same material. We build digital infrastructure on the data of sick people, with tools that learn from that data and return predictions. Each of those words — data, people, learn, predict — holds an ethical decision we would rather make consciously than by default.
We work where getting it wrong has real consequences.
Three territories that rarely share a common standard converge in what we do. Each imposes its own responsibility.
Science
From research we inherit the obligation to claim no more than the evidence allows. A strong hypothesis is not a conclusion, and confusing the two — especially before someone seeking hope — is a form of harm.
Artificial intelligence
A model presents its results with a fluency that can be mistaken for certainty. And it inherits the biases of the data it is trained on: poorly fed, it can perform worse precisely in the populations already underserved.
Clinical data
What a person shares about their illness is intimate. Treating it as mere raw material would betray the trust that makes everything else possible. It is a loan, not a property.
Responsibility is not a statement. It's what we do when no one is watching.
Principles are proven in the hard decisions — the ones that cost time, money or attention. These already guide how we work.
We say only what we can defend
One register for everyone: what we couldn't sustain before a scientific reviewer, we don't claim before an investor either. We promise no cures, no timelines presented as certain. Credibility isn't built by landing the grand claim, but by never having to walk it back.
The tool does not replace judgement
Our models propose; experimental evidence and the professional's judgement decide. When a feature might lead someone to trust it beyond what the evidence supports, we limit it or remove it — however technically attractive and commercially useful.
Data is a trust, not an asset
We do not sell or trade personal clinical data, nor use it for purposes other than those it was given to us for. We understand data protection as the legal expression of a prior duty: not to betray those who entrusted us with something intimate.
We think about who is left out
Four in five people with autoimmune disease are women, and innovation reaches those already better cared for first. We cannot solve access or representativeness today, but we can refuse to deliberately build for a privileged minority.
A work that never ends.
None of these tensions is resolved once and for all. They are resolved every day, in concrete decisions, and they are open to scrutiny: this commitment binds the team, the leadership, the founders and those who invest in the project equally.
We do not claim to be a perfect company, nor to have solved everything we name here. We claim something more demanding: to keep these questions present in every decision, to accept being confronted with our own words, and to correct ourselves when evidence, experience or error teaches us we must.
Behind every principle, a team.
Ethics is not held up by a document, but by the people who make the decisions every day. Get to know who builds Neural Omega, and why.
Meet Neural Omega →