The rapid ascent of Physical AI marks a fundamental transition from digital intelligence that lives behind a screen to embodied systems that interact directly with the physical world. Artificial intelligence was confined to processing text and data to provide answers or generate content. However, the convergence of large multimodal models, high-fidelity sensors, and advanced computing has birthed a new generation of machines, from autonomous transport networks to sophisticated robotics, capable of perceiving, reasoning, and acting within physical environments. This shift represents a massive expansion of the AI market, with some estimates suggesting the physical sector could eventually dwarf the purely digital agent market as machines take over complex real-world workflows.

This transformation is being accelerated by the removal of several historical technical bottlenecks that once kept robotics in the laboratory. The arrival of generative AI and foundation models has granted machines a newfound ability to generalize across different settings, reducing the need for hyper-specific programming for every individual task. Simultaneously, the use of digital twins allows these systems to undergo thousands of hours of training in a safe, parallel environment before they ever touch a physical object. When paired with more resilient hardware and the expansion of IoT connectivity, these advancements allow autonomy to move from simple, repetitive factory movements to nuanced, real-time navigation of public streets and buildings.

Despite this rapid private-sector momentum, a significant gap remains in the strategic readiness of the public sector to manage the economic and security implications of this revolution. While the vast majority of economic and workforce leaders recognize that AI is crucial to their future competitiveness, only a small fraction of states have established a well-defined strategy for responding to its impact. This lack of preparation is particularly risky as AI moves into critical infrastructure, emergency response tools, and transport networks. Unlike a software glitch in a document, a failure in a physical AI system carries serious consequences.

As we move through 2026, the mandate for leadership is to treat Physical AI as a core component of both economic growth and risk management. The intersection of artificial intelligence with physical systems, jobs, and infrastructure is no longer a distant vision but a present reality that demands immediate governance. If state and local leaders fail to create clear frameworks for how these autonomous systems operate within public spaces, they risk being stuck in a perpetual state of reaction, allowing private interests to shape the future of public services.

Build the right capabilities for what’s coming. Book a quick call with the Fractional Source team.

This episode examines the legal and operational implications of Idaho’s House Bill 687, a legislative effort to prohibit state agencies from using artificial intelligence that promotes Diversity, Equity, and Inclusion (DEI) principles. Featuring insights from the Center for Democracy and Technology, the discussion explores how the bill mirrors federal-level executive orders aimed at neutralizing algorithmic bias, while highlighting the immense technical difficulty of defining neutral in a programming context. The conversation concludes that such mandates may create significant procurement hurdles for state agencies, potentially forcing them to abandon modern AI tools altogether to avoid legal non-compliance.

This summit brings together government leaders from across the region to talk about how new technology can make public services work better. It will be held at the JW Marriott in Chicago and will cover topics like cybersecurity, AI, data use, cloud systems, and modern digital services. The event is free for public sector staff, with key attendees including Tom Lynch, Chief Information Officer of Cook County Government. Details →

This webinar brings together government leaders and tech experts to show how AI can pull useful information from messy records like case notes and reports. It highlights Idahoʼs use of an AI tool that reviewed over 2,000 foster care case files in just hours. The session focuses on real-world AI examples, openness, accountability, and new ways agencies can find insights in complex data. Details →

This report shows how fake students and false identities are driving a surge in financial aid fraud. In 2025, nearly $90 million was stolen, with over $40 million going to students who didnʼt exist. The issue comes from disconnected systems, slow manual checks, and small IT teams. The solution is full visibility, real-time monitoring, and AI tools that catch suspicious patterns. Read →

The Trump administrationʼs new AI plan would override many state laws, stopping states from punishing developers for how others misuse AI or from placing heavy restrictions on legal AI use. States would still control where AI infrastructure is built, how agencies use AI, and enforce laws against fraud. Critics say it lacks strong privacy rules and ignores issues like bias and environmental impact.