Human agency
Maintain understandable decision routes, meaningful human oversight and ways to question or reverse consequential outcomes.
Practical test: Can an affected person find the responsible human and seek correction?
Conscious AI Charter · Working framework
Five commitments for technology that serves people.
An invitation to practical dialogue by Tiffany Joy Basse and the emerging Mothers Shaping AI initiative. These are proposed principles to examine and improve, not an established certification or a formal standard.
Purpose
Conscious means intentional, not sentient.
Here, conscious AI means deliberate human attention to the benefits, limits and consequences of AI systems. The word does not claim that the systems themselves are conscious. These commitments encourage people to ask better questions before adopting consequential tools.
Protection of children
Use age-appropriate safeguards, evaluate developmental and emotional impacts, and protect privacy, learning and family agency.
Practical test: Has the system been evaluated for the specific ages and contexts in which it will be used?
Evidence and accountability
Test before deployment, track failures in use, document limitations, and give named people authority to act when harm appears.
Practical test: What evidence would cause the responsible leader to pause or stop?
Public participation
Invite meaningful input from people affected by AI, including communities underrepresented in technical and commercial decisions.
Practical test: Did stakeholder input change a design or deployment decision?
Flourishing without coercion
Support human learning, creativity, relationships and choice, without imposing a single model of success or wellbeing.
Practical test: Can the person decline, challenge or choose a different path without unreasonable penalty?
Put principles into the work
Questions for four kinds of decision-makers.
AI builders
What data and incentives shape this system? How do you test failures, contestable decisions and downstream harms? What will you disclose?
Employers and institutions
Which tasks truly benefit? Who owns deployment, appeal and review? What training, policy and alternatives must come first?
Parents, educators and communities
What will children encounter? What does it normalize about advice, friendship, privacy and authority? How will concerns reach decision-makers?
Policymakers and public agencies
Where are oversight and accountability weakest? How can affected communities participate before a system is adopted?
These questions are exploratory, not legal requirements or a compliance checklist.
Independent reference points
Ground the conversation in evidence.
This charter is Tiffany Joy Basse's proposed set of commitments. The following institutions provide independent guidance; their work is not an endorsement of the charter or initiative.
- NIST AI Risk Management Framework, Core. A voluntary framework organized around Govern, Map, Measure and Manage.
- NIST Generative AI Profile (2024). Additional risks and considerations for generative AI.
- UNICEF, Guidance on AI and Children, version 3.0 (2025). Guidance informed by consultation with children and caregivers.
- UNESCO Recommendation on the Ethics of Artificial Intelligence (adopted 2021). Human rights, dignity, oversight and public participation.