The Knowledge Barrier

Key Statistics

Quick-reference compilation of findings from the ICMA 2024 Survey on Artificial Intelligence in Local Government (n=635) and corroborating research.

Back to The Knowledge Barrier

📋 Survey Sample

B.1

ICMA 2024 Survey Sample

635

Total respondents

78%

CAO / Assistant CAO

67%

Council-manager form

41%

Pop. 5,000–25,000

Source: ICMA (2024)

📈 Adoption & Utilization

B.2

AI Priority Level

• Not a priority48%
• Moderate priority28%
• Low priority20%
• High priority5%

Source: ICMA (2024), q1

B.3

Current AI Utilization by Service Area

• None of the above51%
• Water/Wastewater19%
• Local Policy13%
• Resident Engagement12%
• Energy12%
• Public Safety8%
• Transportation4%
• Built Environment4%

Source: ICMA (2024), q2

B.4

The Opportunity Gap: Current Use vs. Perceived Potential

Current UsePerceived Potential

• Resident Engagement+43 pts
• Public Safety+26 pts
• Local Policy+25 pts
• Water/Wastewater+17 pts
• Energy+14 pts
• Transportation+13 pts
• Built Environment+10 pts

Source: ICMA (2024), q2–q3

🚧 Barriers

B.5

Barriers to AI Adoption

• Lack of AI awareness/understanding77%
• Insufficient trained personnel53%
• No organizational policies/procedures41%
• Insufficient funding40%
• Regulatory/legal compliance concerns22%
• Lack of organizational commitment19%
• Lack of support from elected officials10%

Source: ICMA (2024), q5

⚠ Concerns

B.6

Concerns About AI Implications

• AI-generated disinformation/misinformation70%
• Negative public perception and trust56%
• Potential for increased cyberattacks47%
• Privacy and security of resident data46%
• Digital divide, accessibility, equity20%
• Stakeholder collaboration challenges12%
• Workforce displacement and reskilling12%

Source: ICMA (2024), q6

🏛 Governance

B.7

AI Governance Actions Taken

• None of these actions69%
• Hired/appointed AI strategy staff10%
• Organization-wide AI policy/strategy9%
• AI committee/commission/task force6%
• Organization-wide AI training5%
• Governing body resolution on AI3%
• AI budget line item3%

Source: ICMA (2024), q4

📍 Disparities

B.8

AI Readiness by Community Size

Population	High/Mod Priority	Using AI	Has Policy	Has AI Staff
Under 5,000	15%	40%	2%	2%
5,000–24,999	27%	45%	5%	5%
25,000–99,999	44%	55%	15%	16%
100,000+	57%	73%	25%	27%

Source: ICMA (2024), cross-tabulation by UPOP20cat

B.9

AI Priority by Respondent Role

• IT Director73%
• Chief Information Officer62%
• Chief Administrative Officer29%

Technology leaders are 2.5× more likely to view AI as a priority than the CAOs who set strategic direction.

Source: ICMA (2024), cross-tabulation by q7

B.10

AI Priority by Region

• West38%
• South34%
• Northeast32%
• North Central (Midwest)26%

Source: ICMA (2024), cross-tabulation by UREGN

B.11

Key Ratios

7.7:1

Knowledge barrier vs. political resistance

12.5:1

Large vs. small community AI policy

5.8:1

Disinfo concern vs. displacement concern

2.5:1

IT Director priority vs. CAO priority

1.9:1

AI use with policy vs. without policy

1.8:1

Large vs. small community AI use

Calculated from ICMA (2024) survey data

📚 Corroborating Research

B.12a

Hatz et al. Survey of Local Elected Officials (2025)

• Anticipate increased surveillance from AI83%
• Support government regulation of AI (2023)74%
• Anticipate misinformation from AI69%
• Anticipate threats to data security64%
• Unlikely to make AI decisions soon57%
• Support regulation of AI (2022)56%
• Feel inadequately informed54%

Source: Hatz et al. (2025), PLOS ONE, n=1,028 local elected officials

B.12b

Deloitte State & Local Government AI Survey (2025)

56%

cite lack of gen AI skills as a top-3 scaling challenge

<⅓

of pilots deployed to production (majority of organizations)

Source: Chaurasia et al. (2025), Deloitte Insights

B.12c

MissionSquare AI Survey (2025)

46%

of gov employees report using AI tools

65%

report positive impact on document processing

60%

of employees 35 and under considering leaving

54%

of HR managers expect largest retirement wave soon

Source: MissionSquare Research Institute (2024, 2025)

B.12d

Productivity Research

Study	Key Finding
Noy & Zhang (2023)	40% faster task completion, 18% higher quality with ChatGPT
Dell’Acqua et al. (2023)	40% quality improvement on well-suited tasks; 19 pp decline on poorly suited tasks
Becker et al. (2025)	19% slower with AI tools; perceived 20% faster (40-point perception gap)
Brynjolfsson et al. (2023)	34% improvement for novice workers; small declines for most skilled

Sources: Science, Harvard Business School, METR, NBER

All ICMA percentages rounded to whole numbers. Source: ICMA 2024 Survey (n=635).

© 2026 Alton Henley. The Knowledge Barrier.
altonhenley.com · LinkedIn