ACAS X Is Already in the Cockpit. The AI-and-ATC Debate Is Three Years Too Late. Part 1.

Key Takeaways

• FAA approved ACAS X under TSO C‑219, operational on commercial jets.
• ACAS X uses machine‑learning decision logic, can override ATC instructions.
• SMART rollout slated for Sep 2026, but AI already deployed in flight.
• ICAO codified ACAS X authority, making pilots follow Resolution Advisories.
• Deployment succeeded through 15‑year research, standards, and verification.

Pulse Analysis

The Federal Aviation Administration’s approval of the Airborne Collision Avoidance System X (ACAS X) marks the first large‑scale deployment of machine‑learning‑derived safety logic in commercial aviation. Certified under Technical Standard Order C‑219 and incorporated into ICAO Annex 10 in 2022, ACAS X replaces the deterministic TCAS II with a neural‑network‑compressed decision table that evaluates encounter geometry at sub‑second intervals. When a conflict is detected, the system issues a Resolution Advisory that pilots must follow, even if it contradicts air‑traffic‑control instructions. This last‑resort safety net operates autonomously, providing a real‑time layer of protection that the radar‑based ATC system cannot match.

Despite ACAS X’s proven track record, the public debate has shifted to the FAA’s upcoming Strategic Management of Airspace Routing Trajectories (SMART) program, slated for a September 2026 launch. Media narratives portray SMART as a reckless push to hand control of the skies to AI, yet the regulatory architecture that enabled ACAS X—rigorous standards development, extensive simulation, and formal verification—has already been established. The SMART procurement, involving Palantir, Thales and Air Space Intelligence, risks being judged on a false premise that AI in aviation is a novel, untested concept, ignoring the decade‑long groundwork that made ACAS X possible.

The juxtaposition of ACAS X and SMART underscores a broader lesson for the aerospace sector: successful AI integration hinges on clear authority boundaries, thorough validation, and transparent stakeholder engagement. Regulators should apply the four‑phase framework outlined in recent FAA analyses, defining which ATC functions can be safely augmented and under what failure‑mode scenarios. Airlines and manufacturers that adopt this disciplined approach will reap efficiency gains without sacrificing safety, while policymakers can avoid reactionary headlines. As traffic density rises and unmanned aircraft proliferate, the industry’s next AI milestone will likely follow the ACAS X playbook rather than a rushed, politicized rollout.