Google and American Airlines are testing an AI-powered project to mitigate the climate impact of flights by preventing contrails, the white streaks left by planes. Early results show significant reductions in greenhouse effect.
AI-powered solution targets aviation contrail pollution
Artificial intelligence to reduce aviation contrail pollution. Google and American Airlines are testing the results of a pilot project that uses AI to mitigate the climate impact of flights. The experiment focused on preventing contrails, those thin white streaks that airplanes leave in the sky, which contribute significantly to global warming.
Aviation contrail pollution: The hidden weight of contrails
According to Contrails.org, a non-profit research organization, contrails are responsible for approximately 1-2% of human-caused global warming. When airplanes pass through cold, humid atmospheric zones, soot particles emitted by engines act as condensation nuclei for water vapor, forming ice crystals. These crystals create clouds that trap heat, contributing to the greenhouse effect.
While transitioning to cleaner fuels, such as SAF, represents a long-term solution to reduce emissions, it remains a very costly option. In contrast, a slight modification of flight routes to avoid areas prone to contrail formation would be a more economical and scalable solution in the short term.
AI serving the sky
Google has developed a forecasting tool that identifies areas where contrails are likely to form if an aircraft passes through them. These predictions have been experimentally integrated into American Airlines’ flight planning system, Flightkeys. This allows pilots and flight planners to identify alternative routes or altitude modifications that avoid these critical zones.
The experiment involved 2,400 flights operated by American Airlines between the United States and Europe from January to May 2025. Half of the flights had access to contrail avoidance forecasts, while the other half served as the control group.
The results published on Google’s corporate blog are promising: the 112 flights that adopted route modifications generated 62% fewer contrails compared to the control group. Researchers estimated that this reduced the greenhouse effect associated with these flights by approximately 69%.
Aviation contrail pollution: AI promises a less polluted future
Another interesting aspect that emerged from the test is the substantial invariance in fuel consumption between the two groups, disproving fears that route modifications could significantly increase costs. In a historical phase where airlines struggle to optimize traditional routes due to ongoing conflicts, implementing different planning proves complex; nevertheless, American Airlines intends to continue the collaboration to conduct further studies.
On Google’s part, a broader experimentation is planned in the coming months, along with collaboration with other flight planning software providers to integrate the forecasts into their products. “We know that aviation is one of the most difficult sectors to decarbonize,” stated Dinesh Sanekommu, head of Google’s work on contrails. “The hope is that, both through these AI-based predictions and through these joint operational scientific demonstrations, we can collect useful data and evidence to make the right decisions in the long term.”
