evidoria

← Back to browse

Good practice Imported

Frida — NAV Norway's AI chatbot for welfare and employment service inquiries

Norway · Oslo · See the Norway profile · See the Oslo profile

Evidence: Descriptive / self-reported Top 35% 67/100 · Ask Evidence Copilot about this practice

NAV's AI chatbot Frida handles welfare and employment inquiries 24/7. During the 2020 pandemic surge (250% demand increase), Frida answered 270,000+ inquiries in weeks, resolving 80% without human escalation — equivalent to the capacity of 220 full-time agents.

250%
Demand surge during COVID-19 (Mar-Apr 2020)
>270,000
Inquiries handled in initial weeks of crisis
80%
Resolved without human escalation
220 full-time agents
Equivalent FTE agent capacity
850
NAV contact-centre staff
~2017
In continuous production since

Details

Maturity
Established
Promoter
Norwegian Labour and Welfare Administration (NAV)
Period
2017-present
Keywords
conversational AI, citizen services, welfare, employment, chatbot, NLP, boost.ai

Context

Frida is a conversational AI chatbot deployed by the Norwegian Labour and Welfare Administration (NAV) on the boost.ai platform, operating 24/7 with optional handover to human agents during weekday business hours; it has been in continuous production since approximately 2017 and is described as one of the most mature public-sector conversational AI deployments in the Nordic region.

Objectives

Handle inquiries about unemployment benefits, sick pay, parental leave, pensions, and employment services, freeing NAV's 850 contact-centre staff to focus on complex cases.

Activities

The chatbot fields welfare and employment inquiries with optional handover to human agents; University of Agder (UiA) researchers independently studied NAV's chatbot handover mechanisms.

Results

During the COVID-19 pandemic (March-April 2020), NAV experienced a 250% surge in demand; Frida handled over 270,000 inquiries in the initial weeks, resolving 80% without human escalation — equivalent to the workload capacity of 220 full-time service agents.

Conclusions

The key metrics are reported in boost.ai's vendor case study rather than NAV's official annual reports; public-facing performance transparency and independent auditing could be strengthened, and NAV's 2019 benefits-calculation scandal (a separate matter) underlines the importance of scrutinising AI-adjacent welfare tools.

Implementation

Implementation detail (cost, timeline, staffing, conditions for success) is not yet available for this practice.

Commonly funded by

National / regional programmes Digital Europe Programme

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Similar practices you may find useful