On August 25, autonomous freight company Gatik closed a $200M Series D led by the Qatar Investment Authority and Koch Disruptive Technologies, with Millennium Management, ARK Invest and Intact Private Capital participating. It is the company’s largest round. No valuation was disclosed. Gatik has now raised roughly $500M since it came out of stealth in 2019.
The number worth reading is not the $200M. Co-founder and CEO Gautam Narang named a second one: $600M in contracted revenue.
The number that isn’t a raise
Autonomous vehicle companies have posted plenty of funding totals over the past seven years. Very few post contract totals, for a reason that is easy to guess. Pilots don’t come with long-term contracts.
Gatik is past that. Last year it pulled the safety driver out from behind the wheel on commercial routes. It now runs dozens of fully driverless trucks commercially across several markets, on third-generation hardware that operates around the clock, on surface streets and highways, in light rain and snow.
The customer list is not a pilot list either. Walmart was the first, disclosed back in 2019. Canada’s Loblaw, Kroger and Tyson Foods followed. The largest public deal is with PepsiCo: 41 driverless box trucks moving Frito-Lay product from distribution centres to stores across Dallas, Phoenix and Northwest Arkansas. That multi-year agreement was signed on June 8. The round landed a little over two months later.
41 trucks is a small fleet. 41 trucks against $600M in contracted revenue, 85,000 completed driverless orders and 99% on-time delivery is something else. That is not capital betting on whether the technology arrives. It is capital betting on whether the company can build fast enough to deliver what it already sold.
Gatik took the leg nobody wanted
For seven years, autonomous driving money split roughly four ways: urban robotaxis, long-haul big rigs, sidewalk delivery bots, and closed-site work in mining and construction. Gatik entered none of them. It went after the middle mile, the run from distribution centre to store and store to store.
It started on fixed routes under 10 miles. It now runs dynamic routes with dozens of pick-up and drop-off points, up to 400 miles. In that band it has almost no competition.
The logic of that choice sits on the labor side of the ledger.
A robotaxi faces a consumer, a new origin and a new destination on every trip, and a passenger in the back. A long-haul rig faces thousands of miles of open network and a stack of state-by-state rules. The middle mile touches neither. The origin is your own distribution centre, the destination is your own store, the route barely changes across a year, the freight belongs to the shipper, and someone is standing at the dock to unload. The route geometry is the most stable on the map, the liability chain is the shortest, and the purchasing decision is a line item in a logistics budget rather than a consumer choice.
Which is to say: it is the driving task that is easiest to copy, and the one that requires the least contact with a stranger. Technically and commercially, it was always going to fall first.
Three August stories, read together
North American driverless freight produced three events this month.
On August 14, California issued autonomous truck permits to Kodiak and Aurora, nine days after the Teamsters sued the DMV. On August 20, Nevada cleared 8,000 robotaxis for Las Vegas, with the airport still gated behind a second permit.
The first two are permits. The third is a contract.
Those carry different weight. A permit means the regulator does not object. A contract means somebody has already paid, on a multi-year term. Kodiak and Aurora got the right to operate; Gatik reported $600M of committed revenue behind it. The industry question about driverless freight has been stuck on when it becomes commercial. This round replaces the question. It is no longer whether, it is which year the build slots open.
The Chinese comparison is more direct. JD’s driverless vans now run across more than 20 provinces, with headcount flat, and they run the same leg: own warehouse to own node. Two markets, two regulatory regimes, two capital structures, and the same stretch of road commercialised first in both.
The driver this actually exposes
The American conversation about truck driving jobs defaults to long-haul, the driver who crosses several states and is gone for weeks. That category is under pressure, but it is not the one loosening first.
The middle mile is. Load at the distribution centre before dawn, run 30 to 50 miles to the store, unload, come back, sleep at home. Inside logistics, that job holds a specific position: it is frequently where long-haul drivers go after enough years on the road, trading some pay for not leaving the house. The industry treats it as the back half of a driving career.
That back half now sits exactly on top of the route shape automation eats first. Fixed origin, fixed destination, fixed cadence, the same line run hundreds of times a year. The properties that made it a stable job are the properties that make it reproducible.
Gatik’s own hiring says the rest. The company employs 350 people and plans to use the round to add engineers and operational staff. The operations roles are the genuinely new ones: remote monitoring, exception handling, fleet dispatch, dock coordination. They exist, they are being filled, and they are not the same people or the same cities. A route sheds a handful of drivers who ran a fixed shift; the company adds technical headcount concentrated in a head office and a few operations centres.
One marker is worth watching over the next year. Those 41 PepsiCo trucks cover three regions. If that count reaches three digits, or a second consumer-goods shipper of PepsiCo’s size signs, the middle mile has crossed out of one-customer experiment and into the phase where competitors have to answer it. Logistics procurement is a benchmarking business. Once one shipper books the cost advantage, it shows up on everyone else’s budget agenda.
Sources
- Self-driving truck startup Gatik raises $200M following PepsiCo deal (TechCrunch, August 25, 2026)
- Gatik Raises $200 Million; Series D Led by QIA and KDT as Demand for Driverless Commercial Freight Accelerates (August 25, 2026)
- California let self-driving trucks on the road. The Teamsters sued nine days earlier. (LostJobs, August 14, 2026)
- Nevada cleared 8,000 robotaxis for Las Vegas. The airport needs a second permit. (LostJobs, August 20, 2026)
- JD’s revenue fell. Its profit swung positive. Its driverless fleet went nationwide. (LostJobs, August 13, 2026)
