Artificial Intelligence: Bottlenecks Remain for Scaling Physical AI, But Commercialization Progressing

Research evidence excerpt

Artificial Intelligence: Bottlenecks Remain for Scaling Physical AI, But Commercialization Progressing

Viewpoint |

08 Jul 2026 04:30:00 ET │ 16 pages

Artificial Intelligence

Bottlenecks Remain for Scaling Physical AI, But Commercialization

Progressing

Heath Terry AC

CITI'S TAKE +1-212-723-4624

Citi’s 4th annual Robotics and Physical AI Leadership Conference brought heath.terry@citi.com

together leading founders, investors, and operators at the frontier of AC Andrew Kaplowitzphysical AI to assess where the market stands and the road ahead. The

dominant theme was that commercialization has arrived, but scaling +1-212-816-0642

remains hard. Participants distinguished between the promise of physical andrew.kaplowitz@citi.com

AI models and the operational reality of deploying them in unstructured, Vladimir Bystricky

safety-critical environments. Labor shortages, reshoring, and favorable +1-212-816-6595regulatory tailwinds are accelerating enterprise demand, while data

scarcity, talent constraints, battery limitations, and high deployment costs vladimir.bystricky@citi.com

remain key friction points. Humanoids are generating significant Shelby Spencer

investment enthusiasm, but near-term ROI is being driven by purpose-built +1-212-816-0416

AMRs and specialized systems from companies like Locus Robotics and shelby.spencer@citi.comDexterity. The conference reinforced our view that physical AI is a decade-

long buildout, with durable value accruing to companies that own the data Ashley Kim

flywheel, solve real deployment problems, and meet the highest safety +1-212-816-6689

standards. ashley.kim@citi.com

Industrial Automation — Advances in physical AI are expanding addressable Janna Withrow

markets for automation solutions into new end markets and use cases.