Glossario IA
Il dizionario completo dell'Intelligenza Artificiale
Autonomous system
Artificial intelligence system capable of operating and making decisions without direct human intervention, using its own reasoning processes to achieve predefined goals.
Human supervision
Control mechanism where a human operator can intervene, monitor, or correct the actions of an autonomous system, with varying levels of involvement ranging from passive monitoring to direct intervention.
Critical decision
Algorithmic choice with significant consequences for human life, safety, or fundamental rights, requiring in-depth ethical assessment and clear accountability mechanisms.
Ethical framework
Formal structure defining the principles, rules, and procedures guiding the design and deployment of autonomous systems to ensure their compliance with accepted moral and societal values.
Precautionary principle
Ethical approach requiring that autonomous systems demonstrate a high level of safety and reliability before deployment, particularly when potential risks are serious or irreversible.
Decisional transparency
Ability of an autonomous system to explain and justify its decision-making processes in a way that is understandable to stakeholders, including the criteria used and the underlying reasoning.
Value alignment
Process aimed at ensuring that the goals and behaviors of an autonomous system are consistent with fundamental human values and the intentions of its creators in the long term.
Artificial moral agent
AI system designed to make decisions taking ethical considerations into account, capable of reasoning about moral dilemmas and acting according to predefined moral principles.
Algorithmic ethical dilemma
Situation where an autonomous system must choose between several options, all having negative ethical implications, requiring pre-programmed moral conflict resolution mechanisms.
Ethical verification
Systematic validation process attesting that an autonomous system respects specified ethical constraints and principles, including formal tests and critical scenario simulations.
Ethical auditability
Ability of an autonomous system to be examined and evaluated by independent third parties to verify its compliance with ethical standards and proper decision-making conduct.
Ethical self-regulation
Integrated mechanism allowing an autonomous system to adjust its behavior in real-time to maintain its compliance with ethical principles when faced with unforeseen situations.
Autonomous decision bias
Systematic deviation in an autonomous system's decisions resulting from its learning or programming, potentially leading to discriminations or unfair treatments without corrective supervision.
Ethical hierarchy
Prioritization structure of ethical principles in an autonomous system, defining which values prevail in case of moral conflict and how complex dilemmas should be resolved.
Moral delegation
Transfer of decision-making responsibility from humans to autonomous systems for actions with moral implications, raising questions of legitimacy and ethical responsibility.
Artificial consciousness
Theoretical concept of an autonomous system possessing a form of self-awareness and awareness of its actions, allowing for intrinsic ethical reflection rather than mere rule application.