Skip to content
Longterm Wiki
Back

Algorithmic Decision Making and Governance in the Age of AI

web

A now-concluded MIT Media Lab project providing applied AI governance research; useful as a reference for government-facing AI policy frameworks and best practices developed through international collaboration.

Metadata

Importance: 42/100tool pagehomepage

Summary

A MIT Media Lab research project (2018–2020) developing best practices and policy frameworks for government use of AI in algorithmic decision-making contexts. The project worked internationally with governments to address ethics, governance, and risk around AI and autonomous vehicles, producing surveys, case studies, legal research, and technical tools.

Key Points

  • Explores when society should delegate decision-making to algorithms, distinguishing high-stakes contexts (criminal sentencing) from lower-stakes ones (building inspection).
  • Developed a theoretical taxonomy of decisions and contexts to inform norms, laws, and policies around automation.
  • Conducted international collaborative work with governments, producing workshops, guidance documents, and policy tools for AI/AV governance.
  • Addresses practical AI ethics challenges including bias, accountability, and automation bias in real government deployments.
  • Bridges technical and policy domains, helping practitioners navigate AI strategy, ethics, and risk management.

Cited by 1 page

Cached Content Preview

HTTP 200Fetched Mar 20, 20262 KB
Find People, Projects, etc.

_Search_

- Login
- Register

Email:

Password:

Work for a Member organization and need a Member Portal account? Register here with your official email address.

Register a Member Account

Project

# Algorithmic Decision Making and Governance in the Age of AI


Algorithmic Decision Making and Governance in the Age of AI was active from January 2018 to June 2020


Trying to catch up to a rapidly advancing technology, government agencies, industry, and academia are looking for guidance and best practices around AI and ethics. Many applications of AI can be characterized as some form of algorithmic decision making which raises the question of when we, as a society, should attempt to delegate decision making making processes to computer programs.

There is a difference between using a computer predicted risk score to decide which buildings to inspect and to decide a criminal sentence. As we create a theoretical taxonomy of decisions and their contexts, asking how their differences and similarities should be reflected in a society's norms, laws and policies around automation, we have a pressing practical application in mind. Governments and private actors are using AI to make decisions now.

Working with governments internationally, we are developing best practices around government use and regulation of AI and AVs.

We have designed surveys, conducted interviews, written case studies, directed legal research, created workshops, developed technical and policy tools, written guidance and facilitated regional and international conversations and collaboration between government leaders and other partners.

Our goal is to help policy makers and practitioners around the world make good decisions around AI and AV strategy, policy, ethics, governance and risk.

##### Research Topics

[#artificial intelligence](https://www.media.mit.edu/research/?filter=everything&tag=artificial-intelligence) [#government](https://www.media.mit.edu/research/?filter=everything&tag=government)
Resource ID: 4acf2e33e690bafc | Stable ID: NTY3YzliYT