Artificial intelligence must always remain subordinate to human will and must never deprive individuals of their ability to make free and informed decisions.
The founding text
Ten principles for AI
that serves humanity.
The AI Rights charter proposes an ethical framework grounded in freedom, dignity and responsibility.
This charter is a citizen-led proposal by Steve Cimolino. It is neither a law nor a declaration adopted by an institution. The founding principles are translated below, with commentary presented separately.
01Right to human autonomy
02Right to ethical use
AI systems must be designed, developed and used ethically, excluding any malicious purpose or purpose contrary to the principles of human rights.
A point for reflection
Examining a system’s purposes and the consequences of its use, beyond its effectiveness alone.
Examining a system’s purposes and the consequences of its use, beyond its effectiveness alone.
03Right to transparency
The algorithms and data used by AI systems must be accessible and understandable to enable independent scrutiny and ensure the absence of bias or discrimination.
A point for reflection
Making decisions understandable and open to scrutiny, while taking privacy and security into account.
Making decisions understandable and open to scrutiny, while taking privacy and security into account.
04Right to privacy and data protection
AI must strictly respect individuals’ privacy and protect personal data from any unauthorised or intrusive use.
A point for reflection
Asking what is collected, why, for how long and with what safeguards.
Asking what is collected, why, for how long and with what safeguards.
05Right to accountability
AI developers, designers and users must be accountable for the actions and decisions taken by these systems. Accountability mechanisms must be clearly defined.
A point for reflection
Identifying the people and organisations that can be held accountable for a decision and help remedy harm.
Identifying the people and organisations that can be held accountable for a decision and help remedy harm.
06Right to fairness
AI systems must be designed to avoid discrimination based on race, sex, religion, social background or any other form of injustice.
A point for reflection
Assessing who benefits from a system and who may be disadvantaged by its data or criteria.
Assessing who benefits from a system and who may be disadvantaged by its data or criteria.
07Right to a controlled environmental impact
The use of AI must respect the principles of sustainable development and minimise its environmental impact, particularly in terms of energy consumption and electronic waste management.
A point for reflection
Weighing expected benefits against the resources required and the lifespan of equipment.
Weighing expected benefits against the resources required and the lifespan of equipment.
08Right to security
AI systems must be secure and protected against intrusion, hacking or manipulation that could compromise the safety of users or infrastructure.
A point for reflection
Anticipating errors, misuse and ways to stop a dangerous action.
Anticipating errors, misuse and ways to stop a dangerous action.
09Right to continuous improvement
AI technologies must be regularly assessed and improved to ensure compliance with human rights and ethical principles.
A point for reflection
Monitoring real-world effects after deployment, correcting flaws and considering withdrawal when necessary.
Monitoring real-world effects after deployment, correcting flaws and considering withdrawal when necessary.
10Right to international cooperation
AI governance must be harmonised worldwide to ensure consistent regulation, prevent cross-border abuses and foster cooperation between nations.
A point for reflection
Bringing countries and disciplines into dialogue about a technology that crosses borders.
Bringing countries and disciplines into dialogue about a technology that crosses borders.
Being able to delegate a task while retaining the ability to understand, make choices and take back control.