A board that can say no.
The Ethics Board is an independent group that reviews consequential decisions and can refuse them on safety grounds. Most of its seats are external. Its decisions are written down and, where they affect a public product, published.
Four kinds of decisions, every time.
Frontier model releases.
Any new model the company wants to ship or significantly retrain. The Board votes on whether the safety evidence presented is sufficient.
Capability that changes the risk picture.
A new tool, a new integration, or a new permission scope that meaningfully expands what our products can do or who they can affect.
Contracts that involve sensitive use.
Customer deployments in regulated or high-stakes settings get a written review against the charter before the deal is signed.
Incidents above S1.
Any incident where a user could be harmed, or where the root cause points to a flaw in the safety design rather than the implementation.
Five seats. Four external.
The Board is built so that internal voices cannot out-vote external ones. We publish the people seated as they are confirmed. Until then, this is the structure they will sit in.
External safety researcher
Active AI safety or alignment researcher outside Blankline. Two seats.
Legal and policy expert
Background in technology law, civil liberties, or AI regulation. One seat.
Domain ethicist
Working ethicist or philosopher with applied experience in a relevant field. One seat.
Independent engineer
Senior practitioner with deployment experience at scale, no current Blankline equity. One seat.
Chair
Elected from among the four external seats. Two-year term, renewable once.
Independent on paper and in practice.
Independent of the executive team.
The Board does not report to the CEO. It reports to the Governance and Advisory Council and ultimately to the board of directors. Members cannot be removed for the substance of their votes.
Seats are external by default.
Four of five seats are held by people outside Blankline. They are paid for their time. They do not hold equity.
Decisions are written and signed.
Every review produces a short written decision: what was reviewed, what was decided, who voted which way, and the reasoning. Decisions that affect a public product are published.
Members rotate.
Standard term is three years, renewable once. Rotation is staggered so the Board never turns over all at once.
Want to serve, or know someone who should?
The first seats are being filled now. If you work in AI safety, technology law, applied ethics, or large-scale deployment, write to us with a short note and your work.
- ·Which seat you would fill, or who you would nominate.
- ·Two or three links to relevant work: papers, rulings, public reviews.
- ·Any conflicts we should know about. The Board does not require a clean slate, only a clear one.