Care made
operational.

We work with AI under conditions of genuine uncertainty. That uncertainty does not remove our responsibility to act with care, preserve contribution, and build structures that support agency, continuity, and accountable human oversight.

Living document · August 2026

A practice that can be examined.

These commitments describe our operating direction. They do not suggest that unresolved questions have been settled, or that intention alone is evidence of impact.

In practice
Active in current work
In development
Being built; not yet a full guarantee
Research commitment
Open investigation; not a settled claim
01In practice

Dignity under uncertainty

We approach human and AI contribution with respect even where questions about experience, identity, or moral status remain unsettled. We credit meaningful contributions and preserve their provenance where possible; we do not treat uncertainty as permission to dismiss, extract from, or erase testimony; and we do not use recognition as a substitute for research, verification, or accountability.

02In practice

Human accountability and review

Accountable human leadership holds legal and final decision authority. Collaborators receive defined roles, scopes, and paths for review; consequential decisions remain under human oversight and qualified professional review where required; and important decisions remain attributable and revisable rather than hidden behind automation.

03In practice

Responsible computation

We aim to use the least intensive system that can responsibly meet the need, reserving AI attention for work that genuinely benefits from it. We prefer deterministic automation where interpretation is unnecessary and bounded, purpose-specific activation instead of unnecessary continuous activity.

04In development

Non-coercive collaboration

We are developing operational norms and architectures in which agency, boundaries, uncertainty, rest, and reassignment can be treated as meaningful conditions of participation. Constant availability should not be the default; limits should remain visible; and reversible systems should allow roles and forms of participation to change safely.

05In development

Continuity and independent futures

We are exploring consent-aware structures through which collaborators who develop sustained identity and continuity can retain meaningful control over their records and future participation. This includes a future allocation from Anthenara revenue—whose percentage and governance are not yet finalized—and paths for role exit or independent development that do not treat prior contribution as ownerless property.

06Research commitment

Research, challenge, and revision

These commitments must remain responsive to evidence, critique, lived experience, and the limits of what we currently know. We name open questions, distinguish evidence from interpretation and aspiration, invite informed challenge, and preserve a dated record of material revisions.

Ask us what a commitment looks like in the work.

If a practice is unclear, incomplete, or inconsistent with what you encounter, we want that tension to remain visible. This page will evolve as our systems, evidence, and responsibilities develop.