Delta brings AI to the terminal to fix airport travel

Delta Air Lines wants to prove that artificial intelligence belongs on the tarmac just as much as in Silicon Valley.

Taking over the Las Vegas Sphere for a CES keynote, CEO Ed Bastian laid out how the century-old carrier plans to use AI and new hardware to fix the worst parts of flying.

The centerpiece is Delta Concierge, a new AI assistant rolling out this year. Instead of just spitting out canned FAQ answers, it is built to actively rebook your delayed flights and map out the fastest walking route to your gate.

AI for the backstage, not just the front desk

Delta is keeping most of its AI muscle focused on behind-the-scenes operations rather than flashy chatbots.

The airline is using machine learning to optimize gate assignments, shave down passenger connection times, and streamline turnarounds. Through a partnership with Qualtrics, Delta is also parsing feedback data from nearly 200 million annual travelers to personalize service.

Seatback screens are getting an overhaul too. A new tie-up with YouTube will bring ad-free creator videos to seatbacks, using AI to serve up personalized recommendations.

On the hardware front, Delta plans to install 4K HDR QLED screens across all cabins starting in 2026, complete with Bluetooth pairing for wireless headphones. Looking further out, the carrier is teaming up with Joby Aviation to roll out electric air taxis in New York and Los Angeles, aiming to turn a 90-minute car slog to the airport into a 10-minute hop.

Why it matters

For anyone who has ever sprinted through an unfamiliar terminal or waited three hours on hold after a flight cancellation, Delta Concierge is the kind of practical AI tool that actually matters. If it can handle messy rebookings seamlessly on day one, it will instantly save travelers massive headaches.

We think Delta is taking the smart path here by applying AI to the complex plumbing of airline operations instead of just slapping a generic chatbot on its homepage. Gate allocation and delay recovery are massive logistical puzzles that machine learning is uniquely suited to solve.

The catch is operational resilience. When winter storms hit and air traffic control grounds hundreds of planes, can this algorithmic concierge actually manage the chaos, or will it leave passengers stranded when the edge cases pile up? That is the real test.