Exploring AI Use Cases: Successes and Challenges in the North American Public Sector

Introduction

Artificial Intelligence (AI) is rapidly transforming industries worldwide, and the public sector is no exception. Governments and public agencies across North America are leveraging AI to drive efficiency, enhance decision-making, and improve service delivery. However, implementing AI comes with its own set of challenges and ethical considerations. The AI Used Cases online webinar, featuring experts from various government agencies and organizations, explored the successes, challenges, and future potential of AI within the public sector.

Watch the Full Webinar

It provides in-depth insights into key AI use cases, strategies for overcoming implementation challenges, and discussions on ethics and governance.

Speakers: 

1. Charlie Hamer, Co-Founder, Public Sector Network (Facilitator)
2. Zen Van Loan, AI Data Analytics Implementation Lead, Texas Workforce Commission
3. Nathan Manzotti, Managing Director, Data & Analytics, GSA Center of Excellence
4. Anthony Fisher, Local Agency Security Officer for Data Governance and AI, Colorado Department of Revenue
5. Agata Ciesielski, AI Resident, Department of Homeland Security

Here is a summary of the key takeaways from the session...

1. Identifying and Prioritizing AI Use Cases 

[08:26] Zen Van Loan – Division and AI Data Analytics Implementation Lead, Texas Workforce Commission: 

• Begin with a needs assessment to identify pain points.
• Conduct feasibility and impact studies.
• Start with small-scale projects to ensure manageable risk and scalability.
• Example: Using AI to predict workforce attrition and plan for skill gaps.

[10:06] Agata Ciesielski – AI Resident, Department of Homeland Security: 

• Align AI projects with organizational missions.
• Homeland Security prioritizes critical areas like combating fentanyl trafficking and improving cybersecurity.

[12:24] Anthony Fisher – Local Agency Security Officer, Colorado Department of Revenue: 

• Focus on low-risk, high-impact projects.
• Example: Using AI to detect potholes and improve road maintenance.

[14:59] Nathan Manzotti – Managing Director, GSA Centre of Excellence: 

• Use frameworks like the AI Capability Maturity Model to assess readiness.
• Explore tools such as the GAO AI Accountability Framework for structured evaluation.

2. Successful AI Implementations

[17:51] Agata Ciesielski: 

• Four pillars of success: workforce readiness, innovation pathways, infrastructure, and governance.
• Example: FEMA’s use of computer vision to assess home damage.

[22:08] Zen Van Loan: 

• Texas’ smart city initiatives, such as AI for traffic monitoring and real-time adjustments.
• Automating benefits processing, reducing wait times from 30 days to under an hour.

[27:48] Nathan Manzotti: 

• USDA’s AI prototype for automating beef grading using computer vision.
• National Institute of Standards and Technology’s (NIST) work on detecting child exploitation content.

3. Policy and Governance in AI 

[32:11] Anthony Fisher: 

• Develop human-centric AI policies to ensure accountability.
• Address bias through robust data governance practices.
• Example: A chatbot deployed for public policy queries was fine-tuned to avoid outdated or misleading outputs.

[39:45] Zen Van Loan: 

• Adopt privacy-first design principles.
• Secure internal systems to mitigate data leakage risks.
• Regularly conduct independent audits to ensure compliance.

4. Overcoming Challenges in AI Adoption 

[35:13] Agata Ciesielski: 

• Address barriers around safety, security, and responsibility.
• Example: DHS uses AI to detect suspicious vehicles, enhancing border security while maintaining ethical oversight.

[41:15] Zen Van Loan: 

• Early adopters require education on secure and ethical AI use.
• Provide tools to prevent unintentional data breaches, such as secure AI platforms.

5. Future Trends and Emerging Technologies 

[42:12] Nathan Manzotti: 

• Growth in embedded systems and edge AI capabilities.
• Challenges include limits in data availability for training and reliance on synthetic data.

[44:13] Anthony Fisher: 

• Expect consolidation of niche AI tools into larger platforms by major tech players.
• AI will fill gaps in human resources, enabling faster and more efficient processes.

[46:50] Agata Ciesielski: 

• Increased adoption of agent-based systems and robotics in public services.
• AI interacting with physical systems, such as autonomous vehicles.

Conclusion 

The potential of AI in the public sector is vast, from streamlining operations to addressing critical societal challenges. As discussed during the online webinar, success depends on aligning AI projects with organizational goals, ensuring robust governance, and maintaining ethical standards.