# AI Ethics
The broader moral framework governing AI development and deployment. Distinct from [[AI Safety]] (preventing harm) and [[AI Alignment]] (matching values), AI Ethics encompasses the full spectrum of moral considerations.
Core dimensions:
- **Fairness**: Does the AI treat all users equally? Are outcomes equitable across demographics, or does the system encode and amplify existing biases?
- **Transparency**: Can users understand why an AI made a specific decision? Explainability is not optional; it is a prerequisite for trust.
- **Accountability**: Who is responsible when AI causes harm? The developer, the deployer, the operator, or the user? Clear chains of responsibility must exist before deployment.
- **Consent**: Did users agree to how their data is collected, processed, and used? Informed consent requires genuine understanding, not buried terms of service.
- **Societal impact**: What are the second-order effects? Job displacement, power concentration, surveillance normalization, erosion of human agency.
Every organization deploying AI needs an ethics framework, not just a safety checklist. Safety prevents catastrophic outcomes. Ethics ensures the non-catastrophic outcomes are still acceptable. A system can be safe (it does not crash, does not leak data) and still be unethical (it discriminates, it manipulates, it erodes autonomy).
Connected to but broader than [[Responsible AI]], which operationalizes ethical principles into practices and policies. Ethics asks "should we?"; responsible AI asks "how do we do it right?".
The [[EU AI Act]] represents the first major regulatory attempt to codify AI ethics into law. [[AI Governance]] provides the organizational structures to enforce ethical standards. [[AI Bias]] is one of the most visible ethical failures. [[AI Usage Policy]] translates ethics into concrete organizational rules.
[[Human-in-the-Loop]] design is a practical safeguard, but it is not sufficient on its own. A human approving biased outputs does not make those outputs ethical.
## References
## Related
- [[AI Safety]]
- [[AI Alignment]]
- [[Responsible AI]]
- [[AI Bias]]
- [[AI Governance]]
- [[AI Usage Policy]]
- [[EU AI Act]]
- [[Human-in-the-Loop]]