How agentic, physical and sovereign AI are rewriting the rules of enterprise innovation

Across industries, organizations face the velocity paradox: the pressure to adopt and scale artificial intelligence AI at the enterprise level quickly to stay competitive but also proceed carefully as the technology advances faster than existing operating models can support.

And while AI is already boosting productivity and efficiency, only a subset of organizations is using it to truly reimagine their business rather than optimize what already exists.

Amid this reality, enterprise innovation is gaining momentum in three areas: agentic, physical and sovereign AI. Each unlocks new opportunities and demands new governance approaches and architectural choices.

To understand how these shifts are driving enterprise innovation, Deloitte’s latest report, State of AI in the Enterprise: The Untapped Edge, based on insights from more than 3,200 business and IT leaders around the world, examines how these AI trends are helping organizations move from ambition to activation.

The agentic AI market is poised to reach $45 billion by 2030, up from $8.5 billion in 2026, as organizations look to integrate agentic systems that can plan, reason and execute multistep tasks across business functions.

Use cases are expanding rapidly across industries, with Deloitte’s survey revealing 74% of companies plan to deploy agentic AI within two years. Early wins include customer support, while sectors such as finance, aviation and manufacturing are deploying agents to streamline tasks and augment decision-making.

Strong potential spans supply chain coordination, research and development workflows, knowledge management and cybersecurity.

As adoption grows, companies need strong governance models for risk management, accountability and transparency and to enforce ethical guardrails. Yet only 21% of leaders surveyed currently have a mature governance model for autonomous agents, even though these systems can initiate actions, interface with customers or interact with core business processes.

Companies experiencing the most success are starting with lower-risk applications and building cross-functional governance models to bridge teams across the enterprise – from legal and IT to compliance.

Unlike software-based agentic AI, physical AI brings autonomy into the real world through sensors, controls and robotics, and is rapidly becoming integral to operations worldwide.

Well over half (58%) of respondents to our State of AI in the Enterprise survey said their companies are already using physical AI to some extent and adoption is projected to hit 80% within two years.

While manufacturing, logistics and defence lead the way globally, markets in the Asia Pacific are leading adoption, driving widespread integration of robotics, autonomous vehicles and drones – setting the pace for the next wave of industrial automation.

The practical realities of deploying physical AI – where organizations are navigating equipment investments, safety regulations and ongoing maintenance – illustrate why current deployments have focused on controlled settings such as factories and warehouses.

Robotics play a large role in these environments, with common use cases including collaborative robots on assembly lines, inspection drones with automated response capabilities, robotic picking arms and autonomous forklifts.

Robotics ranks second (20%) among the types of physical AI poised to drive enterprise innovation. Intelligent security systems and/or smart monitoring (21%) and digital twins (19%) are also expected to have long-term impacts on enterprises.

Over time, physical AI will likely become a foundational layer of enterprise transformation, integrated not just into operations but into the strategy that shapes how businesses engage with the world around them.

As organizations scale investments in agentic and physical AI, many are discovering that sovereign AI, where technology is built, matters as much as what it can do and is becoming a key factor in business decisions.

Sovereign AI is about technology ownership and strategic independence. Governments are accelerating investments in digital infrastructure – from AI hardware and software to advanced chips and satellite systems – to help reduce dependence on foreign vendors for important AI capabilities.

At the same time, some governments are investing in or providing incentives for sovereign AI, while others are imposing regulatory requirements on locally owned and operated AI infrastructure. Sovereign AI varies by geography and even sector.

For example, the pressure to source AI technology from foreign vendors varies by region: only 11% of companies in the Americas rely on foreign-sourced solutions for the majority of their AI stack, compared to 32% of companies in Europe, the Middle East and Africa, according to our State of AI in the Enterprise survey.

This heightened focus is prompting enterprises to rethink where AI is developed, how data is stored and employed, which chips are used, and who controls the infrastructure. More than three-quarters (77%) of leaders surveyed in the report said the location of AI development is a key factor when choosing new technologies, while nearly three in five currently build their AI stacks with local vendors.

With almost $100 billion expected to be invested in sovereign AI compute by 2026, sovereign AI has become a strategic challenge for multinational organizations, which must navigate complex requirements that vary by country and create customized solutions for different markets.

As this shift accelerates, companies should consider how AI architecture aligns with sovereignty principles to innovate confidently and operate efficiently across borders.

Ongoing developments in agentic, physical and sovereign AI are expanding the boundaries of what’s possible but successful deployment will depend on an organization’s ability to turn experiments and potential impact into tangible business benefits and a lasting competitive edge.

To actively pursue the shift from AI ambition to advantage, organizations can:

The future belongs to those who orchestrate these capabilities with vision, care and discipline. Don’t wait for the AI landscape to settle; help shape it.