AI Basics
This term refers to a computer system that can perform tasks normally requiring human intelligence.
What is artificial intelligence?
This U.S. agency released the AI Risk Management Framework in 2023.
What is NIST?
This term refers to unfair treatment of groups by an AI system.
What is bias?
This term refers to unintended harmful outcomes from an AI system.
What is an AI risk?
Leaders must do this when communicating AI risks to the public.
What is build trust?
This type of learning uses labeled data to train models.
What is supervised learning?
This major European law regulates high‑risk AI systems.
What is the EU AI Act?
This principle ensures people can understand how an AI system makes decisions.
What is transparency?
This type of attack manipulates model inputs to cause incorrect outputs.
What is an adversarial attack?
When two values conflict, leaders must make this type of decision.
What is a tradeoff?
This is the process of feeding a model new data to improve its performance.
What is training?
These are rules that guide how AI should be developed and deployed responsibly.
What are AI governance principles?
This ethical concern arises when AI systems collect or infer sensitive personal information.
What is privacy?
This term refers to the difficulty of ensuring AI systems behave as intended.
What is the alignment problem?
This type of plan outlines how to respond to AI system failures.
What is a crisis response plan?
This term describes when a model performs well on training data but poorly on new data.
What is overfitting?
This term refers to ensuring AI systems follow laws, standards, and internal policies.
What is compliance?
This term describes the unequal distribution of AI benefits and harms across populations.
What is algorithmic inequality?
This process tests AI systems under extreme or unexpected conditions.
What is red‑teaming?
Leaders use this process to weigh uncertain outcomes when making policy decisions.
What is scenario analysis?
This architecture, introduced in 2017, revolutionized modern AI models.
What is the transformer?
This concept describes the challenge of regulating AI across multiple countries with different values.
What is global governance fragmentation?
This philosophical framework evaluates AI actions based on maximizing overall well‑being.
What is utilitarianism?
This term describes risks that arise when AI systems influence or destabilize entire societies.
What are systemic risks?
This leadership challenge involves coordinating multiple stakeholders with conflicting incentives.
What is multi‑stakeholder governance?