United: explaining a delayed flight with AI, while engineers monitor aircraft health
United has used generative AI to help staff write disruption updates and has consolidated predictive aircraft-health monitoring. The customer communication metrics and maintenance installation tell different stories.
The operating picture
A message to the traveller and a warning to maintenance
Boeing 737-900ER - United Airlines - N37466. Photograph: Oleg Yunakov / CC BY-SA 4.0 (cropped to fit). Illustrates the airline, not an AI deployment.
Passengers can tolerate a delay better when the airline explains what happened and what is likely to happen next. United's AI-supported team writes flight-status messages rather than allowing a model to communicate without oversight.
In maintenance, Lufthansa Technik describes United integrating condition monitoring for Boeing 737 NG and Airbus narrowbody aircraft. The present evidence is clearer about the technology adopted than about how many technical disruptions it has prevented.
Selected implementations
The work behind the evidence
01 / Deployed
Explaining why the flight is delayed
The operational problem
Passengers need accurate, specific information during disruption; misleading explanations damage trust.
What the evidence reports
Forbes described United's AI-supported delay stories in May 2024. Later CIO reporting described more than 100,000 digital notifications, a 6% customer-satisfaction lift and broad use of United GPT among managers.
Operational significance
Human-edited messages are a bounded application where timeliness and factual accuracy can be assessed.
Evidence still needed
The satisfaction change is company-reported; its method and attribution to generated content are not independently established.
Maintenance must judge which component alerts warrant action without grounding aircraft unnecessarily.
What the evidence reports
Lufthansa Technik's April 2022 account describes United co-developing AVIATAR condition monitoring for its 737 NG fleet, connected with Airbus narrowbody monitoring.
Operational significance
An existing engineering workflow can compare alerts with faults found during inspection and actual aircraft availability.
Evidence still needed
The supplier account does not provide current model performance, avoided cancellations or maintenance-cost results.
A source can verify an announcement without verifying its projected outcome.
The wider register
Other tracked activity
Other records are shown separately, including non-AI technology and programmes at different stages of deployment. A register entry does not establish a measured AI benefit.
aiOla voice AI, with investment from United Airlines Ventures · Reported · unverified
United Airlines Ventures took part in aiOla's USD 25 million Series A2 round, aiOla announced on 9 July 2025, and the company said United is exploring its Voice AI ecosystem for a range of potential uses. The platform supports more than 100 languages.
In a Future Travel Experience interview on 10 August 2025, aiOla said United is already using its technology. That claim comes from aiOla.
A source can verify an announcement without verifying its projected outcome.
Editorial assessment
United's customer messaging illustrates how generative AI may assist a human team without owning the final statement to passengers.
Its predictive maintenance work requires the harder test of verified fault anticipation and interventions that genuinely avoided disruption.
Live data from the Aerostack register. Status reflects published evidence only. See Our approach for how records are graded.
Aerostack register · 3 tracked deployments
Aerostack tracks AI deployments across airports and airlines, with the published evidence behind each record. Photography is real aviation imagery. Record thumbnails may show the named airport or airline; other photographs do not.
