The AI landscape doesn't move in one direction — it lurches. Some techniques leap from experiment to table stakes in a single quarter; others stall against regulatory walls, technical ceilings, or organisational inertia that no amount of hype can dislodge. Knowing which is which is the hard part. The State of Play cuts through the noise with a rigorously maintained index of AI techniques across every major business domain — classified by maturity, evidenced by real-world adoption, and updated daily so you always know where you stand relative to the field. Stop guessing. Start knowing.
A daily newsletter distilling the past two weeks of movement in a domain or two — delivered to your inbox while the index updates in the background.
Each dot marks the weighted maturity of practices within a domain — hover for a brief summary, click for more detail
AI systems that navigate and operate in the physical world across road, air, marine, rail, and space. Heavily leading-edge: ADAS and warehouse robotics are good practice, but full autonomy in open environments remains elusive. Evenly split between advancing and stalled — regulatory frameworks and edge-case safety are the binding constraints. One of the largest domains with 22 practices spanning a wide maturity range.
The headline: Autonomy's new gatekeeper is not the regulator — it is the insurer. Underwriters have pulled AI out of standard cover, and the cost is now showing up in bankruptcies and quarterly losses.
Self-driving technology works in the narrow settings it was designed for: mapped robotaxi zones, highway trucking corridors, mining pits, container terminals, rural drone routes, and satellites in orbit. Most operators are past the "does it work" question and stuck on "can we afford to run it." A small group is pulling ahead on volume — Chinese operators dominate robotaxi rides, port automation, mining trucks and sidewalk delivery, and are the only ones whose per-delivery costs beat human labor. Everyone else faces a window in which the technology is available but the economics, the insurance and the public are not yet aligned. If you run fleets, freight, infrastructure or logistics, this year's decision is less about capability than about whether your market will let you use it profitably.
Insurers have written AI out of standard policies — and the first casualties are visible. Industry analysis shows major carriers excluded AI exposure from standard cover from January 2026, citing losses they cannot cap or trace. Air-taxi insurance now runs at 8–15% of aircraft value a year against 1–3% for helicopters — a cost cited as a contributing factor in the collapse of manufacturers Lilium and Volocopter. Before your next renewal, ask your broker in writing whether autonomous and AI-driven operations are covered.
US regulators approved the first cars built with no steering wheel. On 30 July the US road-safety regulator granted Amazon's Zoox an exemption from eight federal vehicle safety standards, allowing up to 2,500 purpose-built driverless vehicles a year. Congress pushed the other way the same week with a bill on how such vehicles must handle emergency responders. Assume federal permission and local restriction keep moving in opposite directions.
The safety case for driverless cars was audited — and it narrowed. An independent study confirmed Waymo's crash-rate advantage overall, but city by city it ranged from a 76% improvement in Phoenix to a 4% increase in Austin. The technology is still safer on the evidence available, but that evidence is thinner than the marketing — which matters if you are leaning on it in a contract or a board paper.
Driverless trucking hit its commercial milestones and its losses at the same time. Aurora logged 440,000 driverless miles with perfect on-time performance and zero collisions, then reported a $270 million quarterly loss on $2 million of revenue. More significantly, TFI International — Canada's largest trucking company — committed to US driverless deployment in 2026–27, the clearest sign yet that established carriers, not just start-ups, are buying in.
Sea drones were used in combat for the first time. US-built uncrewed vessels struck an Iranian naval base on 12 July, and NATO committed $40 billion through 2031 to counter-drone systems across all 32 member states. Defense is the one part of this domain where the cost case is unambiguous — one Marine Corps system engages targets at roughly $10 each versus millions for a conventional interceptor — so military money is now setting the pace of technology everyone else will eventually buy.
China's "hands-off" highway driving becomes a live commercial product. BAIC's Arcfox Alpha S is the first mass-produced car legally approved to drive itself on Chinese highways, and Volkswagen has committed to matching it there by late 2027 — while BMW paused its European equivalent on cost grounds. If you sell vehicles, components or fleet services into China, design to the mandatory 2027 standard now.
Drone delivery is consolidating into a closed club of licensed operators. DoorDash became only the eighth company to win the FAA certification needed to run commercial drone deliveries, joining Amazon, Wing and Zipline. Retail and logistics leaders should settle their drone position in the next two quarters, before the certified few lock up the routes and partnerships.
Space operators face a hard 2028 deadline. The EU now requires every new low-orbit satellite to avoid collisions autonomously by 2028, and the FCC has adopted rules requiring real-time tracking data sharing. Anyone with satellite assets or dependencies should be confirming supplier compliance plans this year, not in 2027.
Working perfectly and losing money are entirely compatible here. Aurora's trucks have a flawless safety record and a $270 million quarterly loss. Sidewalk-robot operator Serve grew revenue 578% to $3 million against $42.8 million of costs, and Hamburg's port operator HHLA cut profit guidance 15–18% over automation disruption. Budget for the operating losses, not just the pilot.
National safety statistics hide local reality. The same fortnight produced the lowest US road fatality rate in twelve years and a 73% jump in crashes involving driver-assistance systems. Whatever the headline number says, your own operating environment is the one that gets tested.
These systems still fail on the ordinary and unexpected. Independent testing found production automatic-braking systems do not reliably detect road workers in reflective safety vests — one model registered no braking at all. Robotaxis still struggle with emergency scenes. The remaining risk sits precisely where headlines, lawsuits and regulators live.
Go deeper: the full Autonomous Systems & Vehicles briefing — the longer analytical write-up, plus every practice we track in this domain with its maturity rating, the tools to consider, and the evidence behind our assessment.