GE Aerospace (NYSE: GE) is stacking alliances with AI-focused defense startups at a pace that signals a deliberate push toward autonomous military aviation. Through separate partnerships with Merlin and Shield AI, the company is coupling its engine and avionics expertise with machine-learning flight software, aiming to build the kind of AI wingmate platforms the Pentagon has been requesting for years. The deals span propulsion hardware, autonomy software, and live flight testing, and they are pulling in major primes like Northrop Grumman along the way.
Autonomy Core Pairs GE Hardware With Merlin Software
The clearest signal of GE Aerospace’s direction came when the company and Merlin jointly announced an autonomy core initiative for advanced aviation. The concept is straightforward but ambitious: merge GE Aerospace’s open system architecture (the standardized digital backbone that already runs across many of its engine and flight management products) with Merlin’s autonomy software stack. The result would be a combined hardware-software package that defense customers could plug into new or existing airframes without starting from scratch on flight control code.
“By teaming with GE Aerospace, we are able to pair their open system architecture with Merlin’s autonomy capabilities to create an important step toward uncrewed flight capabilities,” a Merlin representative said in the announcement. That framing matters because it positions the initiative not as a research exercise but as a near-term integration effort. Open architectures let third-party software ride on top of proven avionics, which shortens the path from lab to cockpit. For GE Aerospace, the arrangement also locks its hardware into the autonomy pipeline early, giving it a structural advantage if the Air Force or other branches scale up orders for AI-piloted aircraft.
Shield AI and the X-BAT Propulsion Deal
Weeks after the Merlin announcement, GE Aerospace disclosed a separate collaboration with Shield AI focused on propulsion technologies for the X-BAT vehicle program. Announced from Cincinnati, the deal targets operations in contested and austere environments, the kind of forward-deployed, communications-degraded scenarios where autonomous aircraft would need to fly without constant human oversight. Shield AI, best known for its Hivemind autonomy software that has already flown on small drones, is working to scale that technology to larger, jet-powered platforms. GE Aerospace’s role centers on the engine side, supplying or adapting propulsion systems that can meet the weight, thrust, and reliability requirements of a combat-rated uncrewed vehicle.
The X-BAT program is distinct from the Autonomy Core work with Merlin in a key respect: it is vehicle-specific rather than architecture-general. Where the Merlin partnership aims to create a reusable autonomy layer for multiple airframes, the Shield AI collaboration ties GE Aerospace’s engines to a particular next-generation platform. Running both tracks simultaneously lets GE hedge its bets. If the Pentagon favors a single large autonomous wingman program, the X-BAT work positions GE as a propulsion supplier. If the services instead adopt a modular, plug-and-play approach across many airframes, the Autonomy Core architecture is ready.
Merlin Expands Into Northrop Grumman’s Test Ecosystem
GE Aerospace’s autonomy partner Merlin is also building credibility with one of the largest defense primes in the country. Merlin signed an agreement with Northrop Grumman Corporation to advance next-generation autonomous flight, joining Northrop Grumman’s test ecosystem for integration and evaluation work. The activities include software-in-the-loop (SIL) testing and flight test operations conducted in Mojave, California, a location long associated with experimental aircraft programs. Merlin’s defense portfolio already includes work on the KC-135, the Air Force’s aging tanker, where autonomy software could eventually handle refueling boom operations or reduce crew requirements on long-endurance missions.
For GE Aerospace, Merlin’s growing list of prime-contractor relationships strengthens the case that the Autonomy Core Initiative is not a paper exercise. When a startup’s software earns a seat inside a Northrop Grumman test environment, it passes a practical credibility threshold that contract officers and program managers watch closely. The Mojave flight test operations, in particular, move the technology out of simulation labs and into real airspace, which is where certification data and operational confidence actually accumulate. If Merlin can demonstrate reliable autonomous performance on legacy platforms like the KC-135 while simultaneously co-developing new architecture with GE Aerospace, the combined offering becomes harder for competitors to replicate quickly.
What the Dual-Track Strategy Means for the AI Wingmate Race
Most coverage of AI wingmate programs focuses on the airframe builders or the autonomy software companies. GE Aerospace’s approach is different: it is positioning itself as the connective tissue between those two camps. Engines and avionics are the systems that every autonomous aircraft needs regardless of who designs the fuselage or writes the flight algorithms. By partnering with both Merlin and Shield AI rather than picking one, GE keeps its technology relevant across competing visions of what an AI wingmate should look like.
This dual-track strategy also aligns with how defense procurement often unfolds in practice. Services rarely commit early to a single architecture or platform; instead, they fund parallel experiments, assess which combinations of hardware and software perform best, and then scale the winners. GE Aerospace’s Autonomy Core work with Merlin is tailored for that experimentation phase, offering a flexible toolkit that can be integrated into multiple demonstrators. The X-BAT propulsion collaboration with Shield AI, by contrast, is a bet on at least one of those demonstrators maturing into a program of record that needs reliable, production-ready engines.
Risks, Timelines, and the Road to Operational Use
Despite the momentum behind these partnerships, the road from test flights to operational squadrons is neither short nor guaranteed. Autonomy in contested airspace raises questions about rules of engagement, cybersecurity, and human oversight that the Pentagon and lawmakers are still working through. Programs that rely heavily on AI decision-making will face intensive scrutiny over how algorithms are trained, validated, and monitored once deployed. GE Aerospace’s emphasis on open system architecture could help mitigate some of these concerns by making it easier for government customers to inspect, update, and certify the autonomy stack over time rather than treating it as a sealed black box.
Technical risk is also significant. Integrating advanced autonomy software with high-performance propulsion systems demands rigorous safety cases, especially when retrofitting legacy aircraft like the KC-135 or pushing new vehicles such as X-BAT into high-threat environments. Flight testing in places like Mojave is designed to surface those edge cases early, but scaling from a handful of test articles to dozens or hundreds of operational aircraft will stress supply chains and certification processes. GE Aerospace’s long history in engine production and sustainment gives it an advantage on the manufacturing side, yet it will still need close coordination with Merlin, Shield AI, and primes like Northrop Grumman to ensure that software updates, hardware changes, and mission requirements evolve in lockstep rather than at cross purposes.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
