TechCrunch Mobility: A robotaxi ultimatum

The Phoenix divorce was inevitable. Uber and Waymo still share robotaxi operations in Atlanta and Austin, but the writing is on the wall: every remaining deal is a countdown. The interesting question isn't when they end. It's what happens the day after.

Uber executives are already firing thinly veiled shots at Waymo. Once the contracts expire, the gloves come off. The battleground shifts from dispatch algorithms to policy — specifically, which markets robotaxi companies are allowed to enter and under what terms. Uber has the lobbyists, the political capital, and the incumbent's instinct to protect its moat. Waymo has the fleet. That collision will define the next five years of urban transport.

Then the federal government walked in. NHTSA administrator Jonathan Morrison issued a directive to every autonomous vehicle developer on the Department of Transportation's Standing General Order. The language was deliberate: "functional insufficiency." Emergency scenes, he wrote, are not rare edge cases. The agency demanded solutions by month's end.

Morrison never named Waymo. He didn't have to. A TechCrunch investigation already documented Waymo's repeated run-ins with first responders across Los Angeles, Phoenix, and San Francisco. Then came July 4. San Francisco gridlock after a fireworks show. Waymo robotaxis sat dead in traffic, batteries drained, blocking lanes while emergency vehicles tried to move. Supervisor Bilal Mahmood is preparing a formal letter of inquiry. The tow trucks did the talking.

This is the moment the industry has been dreading. For years, robotaxi boosters argued that safety data would speak for itself. The data just spoke. It said: your largest fleet cannot handle a predictable traffic jam on a predictable holiday. That isn't an edge case. That is a design failure.

Morrison's letter carries weight. But weight without enforcement is theater. The agency has asked for "solutions" by end of month — a deadline that suggests they expect slideshows, not recalls. The 2026 Regulatory Plan and Unified Agenda, updated last week, tees up a long list of proposed changes to Federal Motor Vehicle Safety Standards. Those standards govern vehicle design and equipment. If NHTSA follows through, the FMVSS rewrite becomes the lever to force architectural changes on robotaxi platforms, not just operational patches.

Waymo will argue its incident rate beats human drivers. That statistic is true and irrelevant. Human drivers don't stall in intersections for hours because their software didn't anticipate a holiday crowd. Human drivers don't require tow trucks to clear a fire lane. The regulatory standard for a commercial robotaxi service is not "better than a tired human." It is "does not obstruct public safety." On that standard, the fleet failed.

Uber watches this closely. Every Waymo stumble is a data point for Uber's policy team. When the Atlanta and Austin agreements dissolve, Uber will argue that Waymo's federal record disqualifies it from new market access. Cities listening to that argument will have cover from the NHTSA letter. The federal rebuke becomes a local veto.

Waymo's countermove is scale. It operates the largest robotaxi fleet in the United States. That scale is supposed to be its moat. But scale without reliability is just a bigger target. The San Francisco tow trucks proved that. The Phoenix exit proved that. The NHTSA letter proved that.

The editorial line is simple: the robotaxi industry has entered its regulatory prove-it phase. No more pilots. No more "safety drivers" as crutches. Morrison gave the industry a deadline. The clock is running. If Waymo and its peers deliver patches instead of architecture, the next directive won't ask for solutions. It will order groundings.

Uber knows this. Its strategy is to survive the transition and own the aftermath. Waymo's strategy is to prove the transition works. Only one of those strategies can win in a market where the regulator just called the leader functionally insufficient. The ultimatum is live.