From the State Government edition of November 3, 2026
North Dakota expands Dakota BOT AI chatbot from the Commerce website to ND.gov after it answered 81% of about 600 questions
North Dakota Department of Commerce and North Dakota Information Technology (NDIT) · State government: digital services and resident information · North Dakota, United States
- Publisher
- State of North Dakota (ND.gov)
- Original publication
- Tuesday, November 3, 2026, 09:45am (ND.gov news)
- Source retrieved
- 2026-11-04
- Event date
- 2026-11-03
What happened
On 3 November North Dakota's Commerce Department and NDIT expanded Dakota BOT, an AI chatbot first launched on the Commerce website in May 2026, to the ND.gov state portal. The state reports that on the Commerce site it fielded nearly 600 questions in more than 400 sessions and answered 81% of them. The bot uses structured workflows and FAQs, was built with input from agencies across state government, and follows an earlier Tourism chatbot. The state says it will evaluate performance as the bot expands to more state websites.
Why it matters
A concrete example of a state scaling a chatbot from one agency site to the statewide portal on the strength of a simple self-reported answer rate, the path many states are considering.
Evidence and measured results
Official state news release, opened in full. Metrics are self-reported; "answered" is not defined and no accuracy, satisfaction or staff-time figures are given.
Limitations and uncertainty
Self-reported, small volume, undefined success metric, no vendor or cost disclosed.
Put this evidence to work
Lighthouse Advisory interpretation, grounded in this source. Enriched 2026-11-04; this does not change the original publication date. Each role follows the same checklist so takeaways compare across sources.
Sales
Role takeaway
- Customer problem
- Residents and businesses struggle to find information across many agency websites, generating repetitive inquiries to staff.
- Stakeholders
- State IT agency
- Commerce or economic development agency
- Portal/web team
- Agency content owners
- Accessibility coordinator
- Discovery questions
- How do you define a successful chatbot answer today?
- Which agencies own the content the bot draws on, and how is it kept current?
- What happens when the bot cannot answer?
- What are current call and email volumes for repetitive questions?
- Value hypothesis
- A curated, workflow-based portal chatbot may deflect some repetitive inquiries; size of benefit is unknown.
- Potential engagement
- Measurement design and content-governance setup for a portal chatbot expansion.
- Unsupported claims
- Accuracy of answers
- Staff time saved
- Cost savings
- Resident satisfaction
Pre-sales engineering
Role takeaway
- Fit
- States with an existing pilot chatbot on one agency site considering a portal-wide rollout.
- Architecture and integration
- Structured workflows and FAQs drawing on agency-provided content, embedded on the state portal.
- Prerequisites
- Curated agency content and FAQs
- Content owners per agency
- Analytics on questions and outcomes
- Escalation to human contact
- Constraints
- Small pilot volume (~600 questions)
- "Answered" metric undefined
- Vendor and model not disclosed
- Security
- Not described in the source; a state would need to define logging, retention and handling of personal data entered by users.
- Proof of value
- Sample transcripts for answer accuracy
- Track unanswered-question rate by topic
- Compare inquiry volumes before and after
Delivery
Role takeaway
- Work
- Expand in stages, define success metrics, set content refresh routines, and evaluate after each expansion.
- Dependencies
- Agency content cooperation
- Portal team capacity
- Analytics access
- Ownership
- Business agency (Commerce) with state IT agency (NDIT).
- Skills and adoption
- Content owners need training on maintaining FAQ and workflow content; staff need a handoff process.
- Governance checkpoints
- Pre-expansion review of accuracy samples
- Accessibility review
- Content currency review
- Acceptance criteria
- Proposed: published definition and rate of answered questions
- Proposed: accuracy sample above an agreed threshold
- Proposed: clear human escalation on every page
- Risks
- Stale content producing wrong answers
- Inflated success metric
- Low usage
Implementation considerations
Lighthouse Advisory interpretation across the operating dimensions a public-sector buyer must settle before this evidence becomes a design. Each note answers the question under its heading for this specific source.
Architecture and integration
What must connect, and where does the AI sit in the workflow?
Workflow- and FAQ-driven design with agency-sourced content suggests a constrained bot rather than open-ended generation; details of model and hosting are not disclosed.
Governance
Who approves, reviews and stays accountable for outcomes?
Expansion decided on an undefined answer-rate metric; the promised ongoing evaluation would need defined measures to be meaningful.
Security and privacy
What data, permissions and controls need testing?
No information on data handling, logging or retention of user inputs.
Accessibility and workforce
Who is affected, and what skills or accommodations follow?
Claims to free staff time from repetitive inquiries; no measurement given. Accessibility and language support not described.
Procurement
What should contracts, pricing and exit terms secure?
Vendor and contract not disclosed.
Operating model
Which teams own the service once it runs?
Joint ownership between a business agency and the state IT agency, with staged expansion across state websites.
Publication history
- 2026-11-03State Government · Issue 571 resources
Stable resource ID: north-dakota-dakota-bot-expands-to-nd-gov-20261103