Fewer technical cancellations and less fuel burned, with AI you can put before your safety office

We price each AI tool against your fuel and maintenance records, then map it to EASA's draft AI rules and to your own aviation authority. We have no aviation client yet, so every result below belongs to an airline or a vendor.

Founded 2025, principal-ledAIGP-certified leadNo junior staff billedRuns on your own cloud

ivory and cloud-grey paper layers traced by one teal line

IATA forecasts jet fuel at 31.4% of airline operating costs in 2026, up from 25.4% in 2025 [1]. SITA says ITA Airways, which runs its AI climb-optimisation tool fleet-wide, is on track to save more than 7,100 tons over 2025 and 2026. That vendor-reported target gives no share of total burn [2].
At 7.8 barrels a ton [3], that fuel is worth $4.98 million at 2025's average of $90 a barrel. At IATA's 2026 forecast of $152, it is worth $8.41 million [1]. SITA gives no split by year, so the true value sits between them, before tool costs. A tool qualifies only if its measured saving on your fleet beats its full cost, and pilots and dispatchers keep authority over every profile.
Sources
1. International Air Transport Association, 'Middle East Disruptions and High Fuel Prices Halve Airline Industry Profitability', IATA Pressroom, Press Release No. 26, 7 June 2026 <https://www.iata.org/en/pressroom/2026-releases/06-07-middle-east-disruptions-high-fuel-prices-halve-airline-industry-profitability/> [accessed 31 October 2026]
2. SITA, 'ITA Airways to Save Fuel and Reduce Emissions in 2026 Thanks to AI-Powered Tool', SITA Pressroom, 23 April 2026 <https://www.sita.aero/about-us/pressroom/news-releases/ita-airways-to-save-fuel-and-reduce-emissions-in-2026-thanks-to-ai-powered-tool/> [accessed 31 October 2026]
3. EUROCONTROL Business Cases, 'Conversions', EUROCONTROL Standard Inputs for Economic Analyses, Release 10.0.2 (Brussels: EUROCONTROL, 2026), Table 4 <https://ansperformance.eu/economics/cba/standard-inputs/latest/chapters/conversions.html> [accessed 31 October 2026]
EUROCONTROL's 2026 standard inputs put an airline's gate delay at €166 a minute averaged over all delay lengths, or €45 for delays under 30 minutes [1]. Both include knock-on delay and are indicative averages in 2022 prices, built on a 2015 study. SITA says airlines using its OCCam recovery tool have cut disruption costs by up to 30%, a vendor-reported ceiling with no airline named [2].
So 1,000 minutes of gate delay cost €166,000. If SITA's ceiling applied to delay cost, which is unproven, a tool would save at most €49,800 per 1,000 minutes. We redesign the control-room process and measure recovery cost per disruption against your baseline before anyone claims a saving. With no published Gulf figure, a Gulf carrier's baseline comes from its own control logs.
Sources
1. EUROCONTROL Business Cases, 'Cost of Delay', EUROCONTROL Standard Inputs for Economic Analyses, Release 10.0.2 (Brussels: EUROCONTROL, 2026), section 16.1, Table 16.1 <https://ansperformance.eu/economics/cba/standard-inputs/latest/chapters/cost_of_delay.html> [accessed 31 October 2026]
2. SITA, 'New Acquisition Brings Proven AI Disruption Recovery to Airlines Worldwide, Reducing Disruption Costs by up to 30%', SITA Pressroom, 1 June 2026 <https://www.sita.aero/about-us/pressroom/news-releases/new-acquisition-brings-ai-disruption-recovery-to-airlines-worldwide/> [accessed 31 October 2026]
easyJet's head of maintenance operations, Aidan Kearney, said predictive maintenance helped avoid 1,343 cancellations between January 2019 and September 2025 [1]. These unaudited figures appear in a 2025 article on an AI vendor's site (vendor-reported); we found no 2026 update. easyJet began reviewing its maintenance approach in 2014 and by late 2025 ran 22 live models on 350 aircraft.
EUROCONTROL's 2026 standard inputs value a same-day cancellation of a 189-seat low-cost flight at €18,570 in 2022 prices [2]. Assuming every avoided cancellation would have been same-day, that gives an upper estimate of €24.9 million, or €3.69 million a year. Once data access is approved, our 30-day proof of concept tests your fault and removal records to pick components worth a model. A licensed engineer makes every release decision.
Sources
1. ePlane AI, 'easyJet Highlights Reliability Improvements Ahead of PAM Dublin 2025', ePlane AI News, 18 November 2025 <https://www.eplaneai.com/news/easyjet-highlights-reliability-improvements-ahead-of-pam-dublin-2025> [accessed 31 October 2026]
2. EUROCONTROL Business Cases, 'Cancellation Cost', EUROCONTROL Standard Inputs for Economic Analyses, Release 10.0.2 (Brussels: EUROCONTROL, 2026), Table 14.1 <https://ansperformance.eu/economics/cba/standard-inputs/latest/chapters/cancellation_cost.html> [accessed 31 October 2026]
Philippine Airlines reports about 30% lower customer-service costs and waits under one minute since moving to the Twilio Flex contact-centre platform (vendor-reported) [1]. AI handles about 10% of tasks, so the figures cannot show what AI alone saved.
In Moffatt v Air Canada (2024), a Canadian tribunal held Air Canada liable for its chatbot's misstatement [2]. Liability stays with the airline even for answers we have tested, so we check them against your fare and refund rules first. Since 2 August 2026, the EU AI Act has required a chatbot's provider to tell users it is AI, unless that is obvious [3][4]. An airline that runs a chatbot under its own name is its provider [5]. Fines reach €15 million or 3% of turnover [3].
Sources
1. Sofiah Nichole Salivio, 'Philippine Airlines Cuts Waits with AI Contact Centre', CMOtech Asia, 25 February 2026 <https://cmotech.asia/story/philippine-airlines-cuts-waits-with-ai-contact-centre> [accessed 31 October 2026]
2. Barry B. Sookman, 'Moffatt v. Air Canada: A Misrepresentation by an AI Chatbot', McCarthy Tétrault TechLex, 19 February 2024 <https://www.mccarthy.ca/en/insights/blogs/techlex/moffatt-v-air-canada-misrepresentation-ai-chatbot> [accessed 31 October 2026]
3. European Commission, 'Quick Facts: Transparency Rules for AI Systems', Shaping Europe's Digital Future, updated 29 July 2026 <https://digital-strategy.ec.europa.eu/en/factpages/quick-facts-transparency-rules-ai-systems> [accessed 31 October 2026]
4. European Commission, 'Article 50: Transparency Obligations for Providers and Deployers of Certain AI Systems', AI Act Service Desk <https://ai-act-service-desk.ec.europa.eu/en/ai-act/article-50> [accessed 31 October 2026]
5. European Commission, 'Article 3: Definitions', AI Act Service Desk <https://ai-act-service-desk.ec.europa.eu/en/ai-act/article-3> [accessed 31 October 2026]
EASA's first AI rule proposal, NPA 2025-07, came out in November 2025, in line with the EU AI Act's high-risk rules [1]. A June 2026 EASA presentation puts the next consultation in Q2 2027, embedding those rules in airworthiness, ATM/ANS and drone regulation [2]. EASA's current AI guidance covers only a 'safe subset': no in-service learning and no part in critical failures [2].
Under EASA's 2018 Basic Regulation, the home state's authority oversees an EU airline or an EU-based Part-145 organisation [3]. A Gulf MRO with EASA Part-145 approval deals with EASA directly [3]. A maintenance model may fall inside those domains, so we map each use case first. We found no GCAA or GACA rule for airline AI, so Gulf carriers should check any file with their regulator.
Sources
1. European Union Aviation Safety Agency, 'EASA's First Regulatory Proposal on Artificial Intelligence for Aviation Is Now Open for Consultation', EASA Newsroom, 10 November 2025 <https://www.easa.europa.eu/newsroom-and-events/news/easas-first-regulatory-proposal-artificial-intelligence-aviation-now-open> [accessed 31 October 2026]
2. Guillaume Soudain, 'EASA AI Roadmap: Implementation Status & Rulemaking Priorities', presentation to the EASA Safety Forum 2026, SKYbrary, 30 June 2026, slides 4 and 5 <https://skybrary.aero/sites/default/files/bookshelf/36207.pdf> [accessed 31 October 2026]
3. European Union Aviation Safety Agency, 'Continuing Airworthiness: General', EASA Frequently Asked Questions, n.d. <https://www.easa.europa.eu/en/the-agency/faqs/continuing-airworthiness-general> [accessed 31 October 2026]

Fewer technical cancellations and less fuel burned, with AI you can put before your safety office

Book a call with Kashif

Kashif Syed

Principal AI Consultant, MENA

IATA forecasts jet fuel at 31.4% of airline operating costs in 2026, up from 25.4% in 2025 [1]. SITA says ITA Airways, which runs its AI climb-optimisation tool fleet-wide, is on track to save more than 7,100 tons over 2025 and 2026. That vendor-reported target gives no share of total burn [2].
At 7.8 barrels a ton [3], that fuel is worth $4.98 million at 2025's average of $90 a barrel. At IATA's 2026 forecast of $152, it is worth $8.41 million [1]. SITA gives no split by year, so the true value sits between them, before tool costs. A tool qualifies only if its measured saving on your fleet beats its full cost, and pilots and dispatchers keep authority over every profile.
Sources
1. International Air Transport Association, 'Middle East Disruptions and High Fuel Prices Halve Airline Industry Profitability', IATA Pressroom, Press Release No. 26, 7 June 2026 <https://www.iata.org/en/pressroom/2026-releases/06-07-middle-east-disruptions-high-fuel-prices-halve-airline-industry-profitability/> [accessed 31 October 2026]
2. SITA, 'ITA Airways to Save Fuel and Reduce Emissions in 2026 Thanks to AI-Powered Tool', SITA Pressroom, 23 April 2026 <https://www.sita.aero/about-us/pressroom/news-releases/ita-airways-to-save-fuel-and-reduce-emissions-in-2026-thanks-to-ai-powered-tool/> [accessed 31 October 2026]
3. EUROCONTROL Business Cases, 'Conversions', EUROCONTROL Standard Inputs for Economic Analyses, Release 10.0.2 (Brussels: EUROCONTROL, 2026), Table 4 <https://ansperformance.eu/economics/cba/standard-inputs/latest/chapters/conversions.html> [accessed 31 October 2026]
EUROCONTROL's 2026 standard inputs put an airline's gate delay at €166 a minute averaged over all delay lengths, or €45 for delays under 30 minutes [1]. Both include knock-on delay and are indicative averages in 2022 prices, built on a 2015 study. SITA says airlines using its OCCam recovery tool have cut disruption costs by up to 30%, a vendor-reported ceiling with no airline named [2].
So 1,000 minutes of gate delay cost €166,000. If SITA's ceiling applied to delay cost, which is unproven, a tool would save at most €49,800 per 1,000 minutes. We redesign the control-room process and measure recovery cost per disruption against your baseline before anyone claims a saving. With no published Gulf figure, a Gulf carrier's baseline comes from its own control logs.
Sources
1. EUROCONTROL Business Cases, 'Cost of Delay', EUROCONTROL Standard Inputs for Economic Analyses, Release 10.0.2 (Brussels: EUROCONTROL, 2026), section 16.1, Table 16.1 <https://ansperformance.eu/economics/cba/standard-inputs/latest/chapters/cost_of_delay.html> [accessed 31 October 2026]
2. SITA, 'New Acquisition Brings Proven AI Disruption Recovery to Airlines Worldwide, Reducing Disruption Costs by up to 30%', SITA Pressroom, 1 June 2026 <https://www.sita.aero/about-us/pressroom/news-releases/new-acquisition-brings-ai-disruption-recovery-to-airlines-worldwide/> [accessed 31 October 2026]
easyJet's head of maintenance operations, Aidan Kearney, said predictive maintenance helped avoid 1,343 cancellations between January 2019 and September 2025 [1]. These unaudited figures appear in a 2025 article on an AI vendor's site (vendor-reported); we found no 2026 update. easyJet began reviewing its maintenance approach in 2014 and by late 2025 ran 22 live models on 350 aircraft.
EUROCONTROL's 2026 standard inputs value a same-day cancellation of a 189-seat low-cost flight at €18,570 in 2022 prices [2]. Assuming every avoided cancellation would have been same-day, that gives an upper estimate of €24.9 million, or €3.69 million a year. Once data access is approved, our 30-day proof of concept tests your fault and removal records to pick components worth a model. A licensed engineer makes every release decision.
Sources
1. ePlane AI, 'easyJet Highlights Reliability Improvements Ahead of PAM Dublin 2025', ePlane AI News, 18 November 2025 <https://www.eplaneai.com/news/easyjet-highlights-reliability-improvements-ahead-of-pam-dublin-2025> [accessed 31 October 2026]
2. EUROCONTROL Business Cases, 'Cancellation Cost', EUROCONTROL Standard Inputs for Economic Analyses, Release 10.0.2 (Brussels: EUROCONTROL, 2026), Table 14.1 <https://ansperformance.eu/economics/cba/standard-inputs/latest/chapters/cancellation_cost.html> [accessed 31 October 2026]
Philippine Airlines reports about 30% lower customer-service costs and waits under one minute since moving to the Twilio Flex contact-centre platform (vendor-reported) [1]. AI handles about 10% of tasks, so the figures cannot show what AI alone saved.
In Moffatt v Air Canada (2024), a Canadian tribunal held Air Canada liable for its chatbot's misstatement [2]. Liability stays with the airline even for answers we have tested, so we check them against your fare and refund rules first. Since 2 August 2026, the EU AI Act has required a chatbot's provider to tell users it is AI, unless that is obvious [3][4]. An airline that runs a chatbot under its own name is its provider [5]. Fines reach €15 million or 3% of turnover [3].
Sources
1. Sofiah Nichole Salivio, 'Philippine Airlines Cuts Waits with AI Contact Centre', CMOtech Asia, 25 February 2026 <https://cmotech.asia/story/philippine-airlines-cuts-waits-with-ai-contact-centre> [accessed 31 October 2026]
2. Barry B. Sookman, 'Moffatt v. Air Canada: A Misrepresentation by an AI Chatbot', McCarthy Tétrault TechLex, 19 February 2024 <https://www.mccarthy.ca/en/insights/blogs/techlex/moffatt-v-air-canada-misrepresentation-ai-chatbot> [accessed 31 October 2026]
3. European Commission, 'Quick Facts: Transparency Rules for AI Systems', Shaping Europe's Digital Future, updated 29 July 2026 <https://digital-strategy.ec.europa.eu/en/factpages/quick-facts-transparency-rules-ai-systems> [accessed 31 October 2026]
4. European Commission, 'Article 50: Transparency Obligations for Providers and Deployers of Certain AI Systems', AI Act Service Desk <https://ai-act-service-desk.ec.europa.eu/en/ai-act/article-50> [accessed 31 October 2026]
5. European Commission, 'Article 3: Definitions', AI Act Service Desk <https://ai-act-service-desk.ec.europa.eu/en/ai-act/article-3> [accessed 31 October 2026]
EASA's first AI rule proposal, NPA 2025-07, came out in November 2025, in line with the EU AI Act's high-risk rules [1]. A June 2026 EASA presentation puts the next consultation in Q2 2027, embedding those rules in airworthiness, ATM/ANS and drone regulation [2]. EASA's current AI guidance covers only a 'safe subset': no in-service learning and no part in critical failures [2].
Under EASA's 2018 Basic Regulation, the home state's authority oversees an EU airline or an EU-based Part-145 organisation [3]. A Gulf MRO with EASA Part-145 approval deals with EASA directly [3]. A maintenance model may fall inside those domains, so we map each use case first. We found no GCAA or GACA rule for airline AI, so Gulf carriers should check any file with their regulator.
Sources
1. European Union Aviation Safety Agency, 'EASA's First Regulatory Proposal on Artificial Intelligence for Aviation Is Now Open for Consultation', EASA Newsroom, 10 November 2025 <https://www.easa.europa.eu/newsroom-and-events/news/easas-first-regulatory-proposal-artificial-intelligence-aviation-now-open> [accessed 31 October 2026]
2. Guillaume Soudain, 'EASA AI Roadmap: Implementation Status & Rulemaking Priorities', presentation to the EASA Safety Forum 2026, SKYbrary, 30 June 2026, slides 4 and 5 <https://skybrary.aero/sites/default/files/bookshelf/36207.pdf> [accessed 31 October 2026]
3. European Union Aviation Safety Agency, 'Continuing Airworthiness: General', EASA Frequently Asked Questions, n.d. <https://www.easa.europa.eu/en/the-agency/faqs/continuing-airworthiness-general> [accessed 31 October 2026]

Book a call with Kashif