A new report from global air travel data provider OAG has warned that the success of AI in aviation hinges on trusted, high-quality information, as poor data inputs risk undermining safety, reliability, and operational performance across the industry.
The OAG Report titled “Why Aviation’s AI Future Hinges on Data Quality” highlights that data integrity is emerging as the critical success factor for the industry’s next technological leap.
The report finds that while airlines are increasingly deploying AI for predictive maintenance, demand forecasting, disruption management, and customer service, most systems remain vulnerable to inaccurate or incomplete datasets.
OAG warns that flawed data, such as outdated flight schedules or inconsistent operational records, could lead to significant downstream consequences, including missed connections, inefficiencies, and financial losses.
According to OAG, AI models are only as reliable as the data they are trained on. In a “zero-defect” industry like aviation, even small inaccuracies can cascade into major disruptions.
For example, inconsistent departure data or outdated minimum connection times can cause itinerary errors, while faulty on-time performance records can distort predictive models for crew or aircraft scheduling.
The report also cites academic and industry research showing that up to 85 percent of enterprise AI projects fail due to data quality issues. In aviation, such failures are amplified by the need for precision, with even minor errors carrying safety and operational risks.
OAG’s analysis calls on airlines and travel companies to make structural investments in data governance and validation. It emphasises that trusted, curated data sources are essential to building the foundation for reliable AI systems. “AI won’t fix aviation’s data problem. Trusted data will,” the report concludes.
The study also identifies emerging AI applications in aviation, from generative itinerary planning to agentic systems capable of autonomous operational decision-making. But it cautions that without consistent, real-time data inputs, these systems risk producing “garbage at the speed of light”.
Follow us on
