Kickoff Meeting · March 25, 2026
Herkimer County AI & Cybersecurity Policy Initiative
A partnership between SUNY Polytechnic Institute and Herkimer County to develop a comprehensive AI and cybersecurity policy framework organized around county departments and operations. Target adoption by the county legislature: November 2026.
1Project scope, deliverables, and timeline
2Shared workspace setup
3Operational materials — scope and protocols
4The seven dimensions — initial conversation
5Workshop staff: identify DSS and department audiences
6Neighboring county outreach — status update
7Next steps and April 22 agenda
2026 Workflow
March to November 2026
All dates are proposed and adjustable. Workshops are 2.5-hour facilitated sessions. Monthly meetings are working sessions with the county administrator, IT director, and DSS leadership.
| Date | Event |
|---|---|
| Wed Mar 25 | Kickoff / Monthly Meeting #1 Today |
| Wed Apr 22 | Monthly Meeting #2: framework development and dimension calibration |
| Wed May 13 | Monthly Meeting #3: synthesis document review and dimension refinement |
| Wed Jun 10 | Workshop #1: AI landscape, values framework, policy options overview Workshop |
| Wed Jul 8 | Monthly Meeting #4: policy draft review |
| Wed Jul 15 | Workshop #2: policy deliberation and staff training track Workshop |
| Wed Aug 5 | Monthly Meeting #5: policy iteration |
| Wed Sep 9 | Monthly Meeting #6: legislative briefing preparation |
| Wed Sep 23 | Workshop #3: policy review and staff training track Workshop |
| Wed Oct 7 | Monthly Meeting #7: legislative briefing preparation |
| Wed Oct 21 | Workshop #4: County Legislature Workshop |
| Wed Nov 4 | Legislative Adoption (target date) |
What We Need from the County
Operational Materials
Digital copies of policy handbooks, employee handbooks, job descriptions, departmental procedures, and forms at whatever scope the county determines is appropriate. These feed the Policy Explorer and ensure policy language reflects how Herkimer County actually operates.
Staff Participation in Validation
Designated windows where county staff review generated policy text against their operational knowledge. Staff build familiarity with the AI and cybersecurity landscape by working through policy options. This serves both quality assurance and training functions.
Workshop Attendance
Participation from county leadership, department heads, and designated staff across the four workshops. Workshops 2 and 3 include focused training tracks organized by department or function as determined by the county.
Neighboring County Outreach
The county administrator has agreed to extend an invitation to peer counties for potential participation. The project methodology is designed for regional transferability, and participation from neighboring counties strengthens the process for all parties.
Discussion: Dimension Calibration, April 22
Monthly Meeting #2 (April 22): This is our first dimension calibration session. We'll work through what each of the seven dimensions actually means for Herkimer County operations.
Core group: County administrator, IT director, and DSS leadership.
A few questions for today:
We'll prepare these materials for ourselves and share them with you in a week or two.
Core group: County administrator, IT director, and DSS leadership.
A few questions for today:
- Should we bring in additional department staff? If so, which ones?
- How would you like to structure their participation?
- Would pre-meeting materials help? We can provide them at three lengths: 25 words, 100 words, or 250 words — covering the full framework plus each dimension separately.
We'll prepare these materials for ourselves and share them with you in a week or two.
The Framing Dimensions
Equity
AI tools and policies are transparent in how they treat employees, residents, service providers, and recipients, including the assumptions embedded in that treatment.
Hi: Equity is an active constraint. All AI tools are reviewed for differential impact before deployment; treatment assumptions are documented and audited.
Neutral: AI tools and policies should not systematically disadvantage any group; differential treatment is unintentional and addressed when identified.
Lo: Equity is a goal but not a procedural requirement; concerns are addressed when raised.
Neutral: AI tools and policies should not systematically disadvantage any group; differential treatment is unintentional and addressed when identified.
Lo: Equity is a goal but not a procedural requirement; concerns are addressed when raised.
Human Impact
The effects of AI-assisted decisions on individuals and families remain visible and subject to human review.
Hi: No AI-assisted decision affecting an individual is final without human review; every determination is documented and appealable.
Neutral: AI-assisted decisions affecting individuals remain subject to human oversight where practicable.
Lo: Human review is available for AI-assisted decisions but not required as a standard step.
Neutral: AI-assisted decisions affecting individuals remain subject to human oversight where practicable.
Lo: Human review is available for AI-assisted decisions but not required as a standard step.
Infrastructure Impact
The material costs of AI adoption — energy use, data infrastructure, procurement, sustainability — are recognized as relevant to county operations and policy.
Hi: Energy use, data infrastructure costs, vendor dependency, and sustainability are formally evaluated as criteria before any AI tool is adopted.
Neutral: Material costs of AI adoption are considered in procurement planning alongside other operational factors.
Lo: Infrastructure costs are tracked but do not constitute a barrier to adoption decisions.
Neutral: Material costs of AI adoption are considered in procurement planning alongside other operational factors.
Lo: Infrastructure costs are tracked but do not constitute a barrier to adoption decisions.
The Operational Dimensions
Security
Protecting county systems, data, and residents from harm or exposure.
Hi: approved, vetted tools; data stays on county networks.
Lo: any available tool permitted.
Lo: any available tool permitted.
Efficiency
Reducing friction and cost in how county work gets done.
Hi: act on AI output directly; no extra approval step.
Lo: supervisor sign-off required before use.
Lo: supervisor sign-off required before use.
Innovation
Encouraging staff to use emerging tools that improve county services.
Hi: staff encouraged to experiment with new tools.
Lo: one county-designated tool only.
Lo: one county-designated tool only.
Accountability
Clear human responsibility for AI-assisted decisions.
Hi: Every use logged with staff ID, tool, timestamp, 90 days.
Lo: No logging required.
Lo: No logging required.
The Policy Explorer
Click a dimension to toggle Hi / Lo. The policy sentence updates instantly. Each dimension's contribution is highlighted in its color.