Skip to content

NCAV and NLI chatbots bring calls from new survivors in Mongolia

ADB's Independent Evaluation Department validated in July 2024 that two AI chatbots in Mongolia brought calls from new users and that 90% of chatbot users, survivors of violence, were satisfied.

  • 2 for a person

    Survivors of violence got information, referrals and counseling through the chatbots, with 90% of users satisfied by figures that the evaluators relay from the completion report.

  • 2 for an organization

    The evaluators found that the chatbot services enabled broader reach to vulnerable, at-risk populations beyond hotlines that could not serve several callers at once.

Depth runs from 1, a convenience or time saved, to 5, a life or a livelihood preserved.

Status
Reported benefit.
Evidence

found, the grade of this benefitstudiedreportedclaimed

An independent evaluation report supports the benefit.

Size

No magnitude yet: frequency is not documented.

Compared with

The technical assistance's outcome target for chatbot user satisfaction. Measured.

Without the system
At least 80% of chatbot users satisfied (target)
With it
90% of chatbot users satisfied
The system’s part
Not isolated. The evaluators credit the chatbot services with the rise in calls from new users, the services provided and the 90% satisfaction, relaying statistics that the completion report took from the project's impact assessment. Neither document separates the chatbots from the live counselors and operators who took many conversations, describes the chatbots' method or shows that they learn from data. The bank's compilation calls the service AI-powered.
Learns from data

Not shown. The evaluators describe a dialogue function built from a range of questions, with 451 unrecognized questions added during operation. Neither document describes a trained model.

“During almost 2 years of operation, 451 unrecognized questions were identified and developed for the chatbot operation.” Source

Later
Persisted, 1 July 2024. The evaluators report that the Ministry of Justice and Home Affairs secured state budget funding to continue the chatbot services and that an ADB project on combating domestic violence supported chatbot counseling in 2023.
Not yet checked
LimitsCosts and harms
Benefit type
Improved Accessibility
System
NCAV and NLI domestic violence chatbots (AI-powered chatbot service on Facebook with mobile texting)
Made by
Anduub Lab LLC
Used by
National Center Against Violence and National Legal Institute, Mongolia, under the Ministry of Justice and Home Affairs
Where
Mongolia
Stage
In routine use
Date
Kind
Aggregate result
Broad benefit
Physical health, Opportunity access (crime reduction or prevention)
Collections
Scrutiny and findings
Last revised
1 November 2026

Sources

  1. Evaluation report, by authors with no tie to the system.

    Asian Development Bank, Independent Evaluation Department. Mongolia: Addressing and Preventing Domestic Violence in Mongolia during the COVID-19 Crisis. Technical Assistance Completion Report Validation Report TCRV-2023-085 (TA 6530, Project 54209-001). Manila, July 2024. Section 2, Evidence of Outcomes Achieved. The validators relay the chatbot statistics from the completion report and the project's impact assessment, so on those figures the document is a descriptive report.

    On the second outcome target, the launch of the two chatbot services by NLI and NCAV led to an increase in calls from new users; and legal advice, referrals, and counseling on family-related issues were provided. Chatbot statistics indicate that 90% of chatbot users, who are survivors of violence, were satisfied to receive online domestic violence services.
    The TA is likely sustainable since the chatbot services enabled broader reach to vulnerable, at-risk populations.
  2. Report, by an evaluator that the makers or users commissioned.

    Asian Development Bank, Information Technology Department. Artificial Intelligence in Action: Selected ADB Initiatives in Asia and the Pacific. Manila, April 2024. Chapter 1, Chatbots as Lifelines for Domestic Violence Survivors, page 9. The financier's own compilation, which relays the technical assistance completion report and the project endline. Copy mirrored by ICTworks in May 2024.

    A total of 26 women and 16 children were rescued because of the chatbot service, while around 30% of the chatbot users were referred for in-depth psychosocial and legal services. The project endline showed that 90% of the users of the chatbot service reported that they were satisfied with receiving online domestic violence support and service.
    Anduub Lab LLC, a local information technology (IT) company, was engaged to develop an AI-powered chatbot service.

Details

AI system

Workflow

Anduub Lab LLC built the chatbots with content drawn from consultations with survivors and frontline workers. Both chatbots ran on Facebook, with a mobile texting function for people in rural areas and people with hearing or speech difficulties. The technical assistance paid for NCAV live counselors and NLI operators who took the conversations that the chatbots did not answer. A national multimedia campaign promoted the chatbots from October 2022.

Follow-up

Conditions for replication

Survivors must be able to reach Facebook or SMS and trust an anonymous channel. Funding for the counselors and the chatbot services must continue, which the Ministry of Justice and Home Affairs secured from the state budget. Unrecognized questions must be reviewed and added to the chatbots' content.

Entry details

Sector

Social services.

Purpose

Anduub Lab LLC built the NCAV and NLI chatbots, under ADB technical assistance, to give domestic violence survivors in Mongolia information, referrals and counseling through Facebook and SMS during the pandemic.

31 more questions not yet checked

model, release, provider, role, configuration, inputs and outputs, autonomy, human oversight, AI contribution, beneficiaries, reach, distribution and access, duration, every measurement question, every money question, ownership and control, dependencies, cross-border connections, geopolitical implications, replication attempts, availability, institutional decision, grounds, action and result, success criterion, every evidence question, observed or intended, alternative assessments, related entries, other identifiers, category mappings.

Cite this benefit

Peter H. Vartanian. (2026). NCAV and NLI chatbots bring calls from new survivors in Mongolia (AIBD-1177). Artificial Intelligence Benefits Database, edition of 2 November 2026. https://petervartanian.xyz/aibd/records/AIBD-1177.html (archived at https://doi.org/10.7910/DVN/AADSII, version 9.0)

BibTeX Data Field definitions