Benefits
- Benefit
-
Raised issues resolved per hour by 15% on average
Awaiting review
- Outcome identifiers
-
Unresolved
Not reviewedUnresolved
- Category
-
Improved Organizational Efficiency
Awaiting review
- Status
-
Reported benefit
Awaiting review
- Beneficiary groups
-
People. Organizations.
Awaiting review
- Beneficiaries
-
Unresolved
Not reviewedUnresolved
- Reach
-
Unresolved
Not reviewedUnresolved
- Distribution and access
-
Unresolved
Not reviewedUnresolved
- Duration
-
Unresolved
Not reviewedUnresolved
- Expected?
-
Yes
Awaiting review
- Costs and harms
-
Unresolved
Not reviewedUnresolved
AI system
- System
-
Generative AI conversational assistant for customer support agents, built on OpenAI's GPT family with additional machine-learning models trained on customer-service conversations
Awaiting review
- Model
-
This is an explicit unresolved assignment: the sources on record do not settle this question.
Not reviewedUnresolved
- Release
-
Unresolved
Not reviewedUnresolved
- Provider
-
Unresolved
Not reviewedUnresolved
- Role
-
Unresolved
Not reviewedUnresolved
- Configuration
-
Unresolved
Not reviewedUnresolved
- Inputs and outputs
-
Unresolved
Not reviewedUnresolved
- Autonomy
-
Unresolved
Not reviewedUnresolved
- Workflow
-
Agents received real-time reply suggestions inside live text chats. The model was trained on a large set of past chats labeled with outcomes. Agents stayed free to ignore the suggestions.
Awaiting review
- Human oversight
-
Unresolved
Not reviewedUnresolved
- Relation to AI
-
The authors compare agents before and after access with agents who had not yet received the tool, in a staggered rollout that began with a small randomized pilot, so the estimate isolates the effect of access to the assistant.
Awaiting review
- AI contribution
-
The authors compare agents before and after access with agents who had not yet received the tool, in a staggered rollout that began with a small randomized pilot, so the estimate isolates the effect of access to the assistant.
Awaiting review
Measurement
- Studies
-
Quasi-experimental. [2]
Fully reviewed: supporting details for Studies
Review date: 29 October 2026.
Sources: [2]
- Study design
-
Quasi-experimental. [2]
Fully reviewed: supporting details for Study design
Review date: 29 October 2026.
Sources: [2]
- Measurement
-
Issues resolved per hour, resolution rate, handle time, net promoter score and attrition from the firm's systems. [2]
Fully reviewed: supporting details for Measurement
Review date: 29 October 2026.
Sources: [2]
- Setting and sample
-
5,179 agents and about 3 million chats in the working paper; 5,172 agents in the published version. [2]
Fully reviewed: supporting details for Setting and sample
Review date: 29 October 2026.
Sources: [2]
- Measures
-
Resolutions per hour. [2]
Fully reviewed: supporting details for Measures
Review date: 29 October 2026.
Sources: [2]
- Comparison
-
Agents not yet given access, with agent and time fixed effects. [2]
Fully reviewed: supporting details for Comparison
Review date: 29 October 2026.
Sources: [2]
- Results
-
14% on average and 34% for novice and low-skilled agents (working paper); 15% on average (published). [2]
Fully reviewed: supporting details for Results
Review date: 29 October 2026.
Sources: [2]
- Additional value from AI
-
About 14% to 15% more resolutions per hour. [2]
Fully reviewed: supporting details for Additional value from AI
Review date: 29 October 2026.
Sources: [2]
- Limitations
-
One firm; staggered rollout chosen by the firm; attrition estimates lack agent fixed effects; small quality declines for the most skilled agents. [2]
Fully reviewed: supporting details for Limitations
Review date: 29 October 2026.
Sources: [2]
Money
- Kind of value
-
The kind of monetary value is unclear.
Not reviewedUnresolved
- Amount
-
The monetary value is not established.
Not reviewedUnresolved
- Currency and prices
-
The currency and price basis are unclear.
Not reviewedUnresolved
- Period and unit
-
The period and unit of the monetary estimate are unclear.
Not reviewedUnresolved
- Realization
-
Whether the monetary value was realized is unclear.
Not reviewedUnresolved
- Recipients and payers
-
The recipients and payers are unclear.
Not reviewedUnresolved
- Gross or net
-
Whether the monetary figure includes costs is unclear.
Not reviewedUnresolved
- Included costs
-
The costs included in the estimate are unclear.
Not reviewedUnresolved
- Comparison
-
The financial comparison is not established.
Not reviewedUnresolved
- Method
-
The method used to calculate monetary value is unclear.
Not reviewedUnresolved
- Return and payback
-
The return on investment or payback period is not established.
Not reviewedUnresolved
People and places
- Participants
-
Developer: An AI customer-service software company that the authors do not name (the 'AI firm'), building on
OpenAI's GPT models. Operator: A Fortune 500 enterprise software company (unnamed) and its outsourced support contractors.Awaiting review
- Jurisdictions
-
Benefit: Support offices mainly in the
Philippines, with smaller groups in the 
United States and other countries. Developer: Not stated. Operator: United States.Awaiting review
- Ownership and control
-
Unresolved
Not reviewedUnresolved
- Dependencies
-
Unresolved
Not reviewedUnresolved
- Cross-border connections
-
Unresolved
Not reviewedUnresolved
- Geopolitical implications
-
The geopolitical implications are unclear.
Not reviewedUnresolved
Follow-up
- Conditions for replication
-
The work must consist of repeated text conversations on recurring problems. The firm must hold enough past conversations to train the model. Gains depend on a supply of less experienced agents.
Awaiting review
- Replication attempts
-
1 follow-up entry. Benefit recurred: 1.
Awaiting review
- Later findings
-
Unresolved
Not reviewedUnresolved
- Availability
-
Unresolved
Not reviewedUnresolved
- Institutional decision
-
A relevant institutional decision has not been established.
Not reviewedUnresolved
- Grounds
-
The grounds for an institutional decision are unclear.
Not reviewedUnresolved
- Action and result
-
The result of an institutional decision is not established.
Not reviewedUnresolved
- Success criterion
-
The criterion for a successful institutional decision is unclear.
Not reviewedUnresolved
Evidence
- Source numbers
-
AIBD-0664-source-0001–0002
Awaiting review
- Citations
-
Numbered references below.
Awaiting review
- Language
-
Unresolved
Not reviewedUnresolved
- Source relationships
-
Independent
Awaiting review
- Original evidence
-
Unresolved
Not reviewedUnresolved
- Access and versions
-
Unresolved
Not reviewedUnresolved
- Supported findings
-
Unresolved
Not reviewedUnresolved
- Evidence links
-
Unresolved
Not reviewedUnresolved
- Evidence strength
-
Independently studied. Sources: one independent peer-reviewed study and one independent preprint. The cited evidence includes no finding by a regulator or court. [1] [2]
- Missing evidence
-
Unresolved
Not reviewedUnresolved
Grades
- Assessed benefits
-
People. Organizations.
Awaiting review
- Depth
-
People: 2. Organizations: 2.
Awaiting review
- Reasons
-
People: Novice and low-skilled agents resolved 34% more issues per hour in the working paper estimate, and attrition among agents with under six months of tenure fell by about 10 percentage points. Organizations: The firm's agents resolved 15% more issues per hour on average after gaining access.
Awaiting review
- Observed or intended
-
Unresolved
Not reviewedUnresolved
- Rubric
-
Unresolved
Not reviewedUnresolved
- Assessor and date
-
Unresolved
Not reviewedUnresolved
- Alternative assessments
-
Unresolved
Not reviewedUnresolved
- Fill
-
Independently studied
Awaiting review
Entry details
- Kind
-
The case describes an evaluation or a study. An academic study of a workplace rollout. [1]
Fully reviewed: supporting details for Kind
Review date: 29 October 2026.
Sources: [1]
- Related entries
-
Both cases concern generative AI that writes replies in a company's customer service chats, where faulty replies fall on the company that provides the support.
Fully reviewed: supporting details for Related entries
Compared with the incidents in the AI Incident Database snapshot of 26 October 2026; a second reading confirmed that the incident concerns the same kind of system and stakeholder.
Sources: incidentdatabase.ai
- Field
-
Customer service.
Awaiting review
- Purpose
-
The assistant was built to suggest responses to customer support agents during live chats. Faster and better resolution of customer issues was that purpose.
Awaiting review
- Dates
-
Source information as of: November 2020. Updated: 29 October 2026.
Awaiting review
- Stage
-
The system was in routine use. The firm rolled the tool out to its own and outsourced agents after a pilot. [1]
Fully reviewed: supporting details for Stage
Review date: 29 October 2026.
Sources: [1]
- Version
-
SHA-256: 9ef1083d747e…
Awaiting review: supporting details for Version
SHA-256 of the public case, its assessments, and its displayed answers and notes, excluding its own identifier and the derived coverage summary.
References
- Other identifiers
-
AI Incident Database: 631.
Awaiting review
- Category mappings
-
Unresolved
Not reviewedUnresolved
- Citable summary
-
Unresolved
Not reviewedUnresolved
- Revisions
-
2 revisions. Latest: 29 October 2026.
Awaiting review
Sources
-
Peer-reviewed study Independent author
Brynjolfsson E, Li D, Raymond L. Generative AI at Work. Quarterly Journal of Economics. 2025;140(2):889-942. doi:10.1093/qje/qjae044 (abstract as listed on RePEc IDEAS).
Access to AI assistance increases worker productivity, as measured by issues resolved per hour, by 15% on average, with substantial heterogeneity across workers.
-
Preprint Independent author
Access to the tool increases productivity, as measured by issues resolved per hour, by 14% on average, including a 34% improvement for novice and low-skilled workers but with minimal impact on experienced and highly skilled workers.