Skip Navigation
Austin, United States

Explore and establish a process for AI accountability in the City of Austin (USAU0007)

Overview

At-a-Glance

Action Plan: Action plan – Austin, United States, 2024 – 2028

Inception Report: Not available

Commitment Start: Jan 2025
Commitment End: Nov 2025

Institutions involved:

  • City of Austin- Office of Innovation
  • City of Austin - Equity & Inclusion
  • Austin AI Alliance
  • University of Texas Good Systems
  • University of Texas IC² Institute

Primary Policy Area:

Primary Sector:

OGP Value:

  • Technology and Innovation for Transparency and Accountability

Description

Commitment ID

USAU0007

Commitment Title

Explore and establish a process for AI accountability in the City of Austin

Problem

Artificial Intelligence (AI) tools are rapidly developing and being adopted by government organizations for a range of tasks and services. Many communicate directly with and/or influence decisions and outcomes for residents. However, due to their design, these tools can generate biased or incorrect outputs and could be misused or abused. Currently, governments lack a strong, systematic framework for identifying, assessing, and addressing bias, errors, and abuse in AI tools.

Status quo

Today’s AI tools' tendency to produce biased and incorrect information is well known; however, solutions to address this are inconsistent. They range from assigning accountability to the software developer to performing testing and monitoring within the organization, to accepting bias and errors as a part of currently available AI tools and mitigating situations as they occur. There have been multiple instances of harmful outcomes for residents due to bias and errors from government-employed AI tools, yet there is no agreed-upon approach for proactively assessing, monitoring, identifying, and addressing them.

Action

The City of Austin will co-create, test, and deploy a framework for AI accountability in the organization. This includes shared processes for assessing AI tools under consideration for purchase and use to understand their structure, potential for bias, and mitigation measures. The framework will also create structures for City staff and residents to detect and report instances of AI errors, bias, and abuse, and procedures to investigate and address these cases. The intention is to ensure safety and accuracy in AI tools the City employs, and that they reflect the value of the Austin community and build trust between government and citizens.

How will the commitment contribute to solving the public problem described above?

Co-creating a standard and shared framework for AI accountability builds a foundation of trust and a culture of proactive prevention when it comes to AI bias, errors, and abuse. Both the City organization and the Austin community see the potential of the rapidly advancing AI ecosystem, but also understand the risks and shortcomings that current tools carry. By creating a framework for assigning accountability, mitigating risk, and addressing challenges, we create a strong base for safe, ethical, and effective AI development and adoption in the future.

What long-term goal as identified in your Open Government Strategy does this commitment relate to?

In the long term, we want to advance the City’s capacity to successfully assess and adopt emerging technologies in a safe, equitable, and ethical manner. Building a framework that is based on trust, accountability, and engagement with the community is key to doing so. Accountability for AI in government is also at the forefront of national and international conversations, and this work will help materially advance the practice there and create an opportunity for others to adapt the framework for use in their jurisdictions.

Primary Policy Area

Digital Governance

Primary Sector

Science & Technology

What OGP value is this commitment relevant to?

Technology and Innovation for Transparency and Accountability Establishing AI accountability in Austin ensures transparency through open documentation and explainable decisions while fostering trust with oversight that upholds ethical and equitable standards.

Milestones

3 Milestones
1

Create project coalition

Start Date12/2024
End Date02/2025
  • Not started
  • In progress
  • Stuck
  • Finished
  • Incomplete
2

Develop, test, and iterate accountability framework

Start Date01/2025
End Date10/2025
  • Not started
  • In progress
  • Stuck
  • Finished
  • Incomplete
3

Deliver and deploy AI accountability framework in City operations

Start Date09/2025
End Date12/2025
  • Not started
  • In progress
  • Stuck
  • Finished
  • Incomplete


Commitments

Open Government Partnership