GE Aerospace Teams Up with Merlin to Build AI Flight Systems for Air Force Tankers

The Sky’s New Autopilot: GE and Merlin Aim to Redefine Aviation

Inside a GE Aerospace lab in Cincinnati, engineers are wiring up something that could change the way planes fly. Think of it as a cockpit’s operating system—a digital brain designed to take on much of the heavy lifting from pilots. Today, GE Aerospace announced it’s teaming up with Boston-based Merlin to make that vision real, starting with the U.S. Air Force’s fleet of KC-135 refueling tankers.

This isn’t just another defense industry handshake. It’s a sign that aviation is steadily moving into an era where artificial intelligence doesn’t just assist pilots—it shares the cockpit with them.

Old Planes, New Brains

The timing makes sense. The Air Force still relies on KC-135 tankers that first flew back when Eisenhower was president. Their cockpit parts are increasingly rare and expensive to replace. To keep the planes in service, the Air Force launched the “Center Console Refresh” program, which also created an opening for advanced technology.

GE already has a foot in the door: its Flight Management System runs on more than 14,000 aircraft worldwide. By combining that proven system with Merlin’s autonomy software, both companies believe they can transform how these tankers operate.

Matt Burns, who oversees Avionics Systems at GE Aerospace, says the collaboration blends “proven expertise with modular design and Merlin’s autonomy know-how.” Merlin’s CEO Matt George is more blunt. For him, this is about national security. “Integrated autonomy isn’t optional,” he argues. “It’s essential if we want to stay ahead.”

How the “Autonomy Core” Works

What GE and Merlin are building isn’t a wholesale cockpit replacement. Instead, the “autonomy core” sits on top of existing flight systems like an extra layer of intelligence. The goal is simple: let the AI handle repetitive, routine tasks while pilots focus on mission-critical decisions.

Flight autonomy is categorized into distinct 'levels,' representing a spectrum of capabilities. These range from advanced pilot assistance systems that augment human decision-making to fully autonomous flight where the aircraft operates with minimal or no human intervention.

The design follows Pentagon rules for modular, open systems. That means it can work across different aircraft without locking the military into a single vendor. Over time, the software could manage navigation, optimize flight paths, and even take part in tricky aerial refueling missions. Machine-learning algorithms would crunch sensor data on the fly, adjusting to weather, turbulence, or enemy threats.

Why the Push Toward Autonomy

There’s no shortage of reasons. Militaries and airlines alike face pilot shortages. Tanker missions often stretch over 12 hours with multiple crew members, sometimes in stormy skies or hostile regions. Cutting down on crew demands could make missions safer, cheaper, and easier to schedule.

Projected pilot shortages in major global regions highlight a key driver for developing autonomous flight technologies.

Region	Projected New Pilots Needed (2025-2044)	Source
Global	660,000	Boeing 2025 Pilot and Technician Outlook
Eurasia	149,000	Boeing 2025 Pilot and Technician Outlook
China	124,000	Boeing 2025 Pilot and Technician Outlook
North America	119,000	Boeing 2025 Pilot and Technician Outlook
Middle East	67,000	Boeing 2025 Pilot and Technician Outlook
Southeast Asia	62,000	Boeing 2025 Pilot and Technician Outlook
South Asia	45,000	Boeing 2025 Pilot and Technician Outlook
Latin America	37,000	Boeing 2025 Pilot and Technician Outlook
Africa	23,000	Boeing 2025 Pilot and Technician Outlook
Northeast Asia	23,000	Boeing 2025 Pilot and Technician Outlook
Oceania	11,000	Boeing 2025 Pilot and Technician Outlook

The aviation industry is also shifting toward software-driven systems. Giants like Honeywell, RTX, and Northrop Grumman are chasing similar autonomy projects. The consensus is clear: smart flight systems aren’t a futuristic dream anymore. They’re the next big leap.

First the Military, Then the Rest

History shows the military usually tests technologies long before the public sees them. This partnership gives autonomy a proving ground where regulators are more flexible and where safety standards differ from passenger aviation.

Civilian regulators are more cautious. For instance, Europe’s safety agency recently paused some single-pilot research programs, saying the risks weren’t fully addressed. That makes the military a safer bet for early adoption.

Cargo flights may come next. Unlike passenger jets, cargo planes face fewer public acceptance hurdles. Analysts predict freight operators will experiment with autonomy well before commercial airlines seat paying passengers behind an AI-assisted cockpit.

Business Stakes

For GE, this partnership isn’t just about tech—it’s about money. Its massive installed base could generate steady revenue from software updates and new features. Instead of selling hardware once, GE could turn avionics into a subscription-style business.

Merlin gains scale and credibility it couldn’t achieve alone. Tapping into GE’s global network could accelerate its path to public markets, especially since the company has hinted at going public via a SPAC deal. If KC-135 upgrades succeed, the model could extend to C-130s, maritime patrol aircraft, and other platforms where reducing crew saves resources.

The Challenges Ahead

Of course, no one should assume this will be smooth sailing. Cybersecurity looms large—if AI flies planes, hackers and electronic warfare become bigger threats. Regulators also need to figure out how to certify AI flight systems, which is far from straightforward.

Certifying AI for aviation safety presents significant challenges, particularly given the rigorous FAA AI certification process. A core difficulty lies in developing explainable AI for aerospace, essential for understanding and ensuring the safety and reliability of AI decisions in critical flight applications.

Then there’s the human side. Pilots still need manual flying skills, even if the AI does most of the work. Training programs will have to strike a balance between man and machine.

The Road Ahead

Don’t expect a sudden revolution. Aviation rarely works that way. Instead, autonomy will likely appear first in small, manageable steps—automated taxiing, smoother cruising, or standard landings. Only later will it tackle complex missions like aerial refueling or combat support.

If GE and Merlin can prove the tech on KC-135s, it could open the door for broader adoption. Cargo carriers may follow, with passenger jets trailing years later once safety rules and public trust catch up.

For now, the GE-Merlin partnership looks like one of the most significant steps yet in blending human judgment with silicon smarts. The cockpit of the future might not replace the pilot, but it will certainly share the workload—and change the way we think about flying.

House Investment Thesis

Category	Summary Analysis
Overall Thesis	The GE×Merlin partnership is a credible entry into a structural retrofit/autonomy cycle for defense mobility fleets (e.g., KC-135, C-130J). The near-term opportunity is MOSA-compliant avionics refresh + "pilot-assist" autonomy, not civil passenger Single-Pilot Operations (SPO). The winner will control the runtime, certification, and installed base.
Why Now? (Drivers)	1. Aging Fleets: Mandates like the KC-135 Cockpit Connectivity Retrofit (CCR) create funded touchpoints for new compute that can host autonomy. 2. Regulatory Pull: DoD's hardened MOSA requirement favors modular, open systems. 3. Crew Constraints: Pilot shortages and contested logistics make reduced-crew operations financially attractive in defense. 4. Installed Base: GE's 14k+ Flight Management System (FMS) deliveries provide a leverageable base for high-margin software attach.
Market Validation (Trend)	Yes. Key players moving: • Honeywell×Merlin: Validates the "autonomy in existing cockpits" approach. • Northrop's "Beacon": Proof that primes want to broker autonomy ecosystems. • Reliable Robotics/Xwing: Operational beachhead for autonomous defense cargo. • Joby buys Xwing: Air taxi OEMs acquiring autonomy IP. • RTX/Collins: Signaling a platform fight for open flight-deck automation. Counter-Signal: EASA has paused civil SPO research, pushing near-term focus to defense + cargo.
GE×Merlin Advantage	• Distribution Moat: GE's FMS penetration and the CCR program provide an immediate host and route to fielding. • Certification Alignment: GE's modular FMS aligns with DoD's MOSA rulebook, lowering procurement friction. • Real Glidepath: Merlin has a CRADA with the USAF, an accepted airworthiness plan, and delivered milestones.
Investment Implications	• GE Aerospace (Positive Skew): Thesis is attaching high-margin software/retrofit kits to its installed base via MOSA programs (CCR first). Catalysts: CCR award, flight demos, civil cargo evaluation. • Honeywell/RTX (Competitive): Well-positioned as alternative "autonomy cores." Expect multi-winner market; watch for first multi-year retrofit win. • Autonomy Vendors (Selective): Defense will fund mission-bounded autonomy first. Best route to scale is via OEM/avionics distribution (e.g., Northrop Beacon). Merlin's SPAC carries execution risk.
Key Metrics to Track	1. CCR Milestones: RFP, award, and production. 2. USAF Flight Tests: KC-135 demo cadence. 3. Regulatory Posture: FAA/EASA movement on cargo (not passenger) autonomy. 4. MOSA Conformance: Adoption in actual awards. 5. Ecosystem Momentum: Northrop Beacon partner adds, competitor demo flights.
Scenarios (Probabilities)	• Base Case: Autonomy-assist scales on 1-2 defense fleets by 2027; CCR awarded; civil cargo evaluations begin. • Bull Case: Rapid CCR execution leads to follow-on buys; a U.S. cargo operator announces an evaluation with GE. • Bear Case: CCR schedule slips to a rival; autonomy stalls on certification/cyber concerns; EASA stance chills valuations.
Sharp Takes	• "FMS is the platform." The gatekeeper for cockpit autonomy will control FMS + MOSA backplane + certification. • "CCR is an autonomy Trojan horse." Framed as sustainment, it pre-wires jets for autonomy. • "Civil SPO is a mirage (for now)." EASA's pause removes near-term TAM; focus on defense/cargo.
Positioning Logic	• Core: System integrators with installed base and certification muscle (GE, HON, RTX). • Selective Beta: Merlin for torque to defense autonomy, sized for SPAC/execution risk. • Theme Basket: Tools for autonomy assurance (certification, cyber-hardened compute). • Avoid: Building models on civil SPO before 2030.

NOT INVESTMENT ADVICE