The next challenge for enterprise AI may not be building more capable models. It may be figuring out how much authority companies should give those models once they become capable of acting on their own.

That question will sit at the center of the 2026 ISG AI Impact Summit London, where Information Services Group (ISG) plans to bring executives from organizations including Lloyds Banking Group, NatWest, Ogilvy, AstraZeneca, Diageo, Reckitt, Carlsberg and Shell together to discuss the move toward what it calls the autonomous enterprise.

The September 9–10 event at Park Plaza Victoria London will focus on practical approaches to redesigning operating models, workforce strategies and data foundations around AI. The agenda suggests that enterprise AI is increasingly being treated as an organizational transformation issue rather than simply a technology deployment.

Eleanor Matthews, director at ISG and host of the summit, said companies are redesigning how work, decision-making and accountability operate as they move toward AI-first models.

That distinction is increasingly important. Generative AI can improve knowledge work, but agentic AI systems introduce a different level of operational risk because they can potentially plan tasks, interact with enterprise systems and execute workflows with limited human intervention.

From AI Pilots to Measurable Business Value

One of the summit's central discussions, “From Pilot to Payback: Closing the AI ROI and Maturity Gap,” will examine why large AI investments do not necessarily translate into measurable business outcomes.

Executives from Lloyds Banking Group, NatWest and Ogilvy will discuss the gap between AI investment, organizational readiness and return on investment.

The problem is familiar across enterprise technology. Companies can demonstrate that an AI model works in a controlled environment without proving that it improves a business process at production scale. Data quality, integration costs, security controls, employee adoption and workflow redesign can all determine whether an AI initiative produces meaningful returns.

For CIOs and CTOs, that shifts the question from “Where can we use AI?” to “Which business processes should be redesigned around AI?”

The distinction also affects marketing and customer-facing organizations. Marketing automation, customer data platforms and AI-powered analytics are increasingly moving toward autonomous decision support, but their value depends on the quality of the underlying data and the organization's ability to govern automated actions.

Sovereign AI Moves Into the Enterprise Conversation

Another major theme will be sovereign AI, reflecting growing concern over where enterprise data is processed, who controls AI infrastructure and how regulatory requirements affect the deployment of external models.

The “Sovereign AI: Who Actually Owns Your Intelligence?” panel will feature James Gill of Lewis Silkin LLP and Oliver Patel, head of Enterprise AI Governance at AstraZeneca.

For multinational businesses, AI sovereignty has become intertwined with cybersecurity, intellectual property, data residency and regulatory compliance. Enterprises operating across Europe and other jurisdictions increasingly need to understand how sensitive data moves between cloud providers, AI platforms and third-party applications.

The issue becomes even more complicated with agentic AI. An autonomous system may access multiple databases, call external services and trigger business processes. That creates a larger governance surface than a conventional software application.

Microsoft, Amazon and Google are competing to provide the cloud and AI infrastructure underpinning many of these enterprise deployments, while companies such as Salesforce and Adobe are embedding AI agents into customer-facing workflows. The resulting market is moving toward AI systems that are increasingly connected to enterprise data and operational applications.

Data Readiness Could Become the Limiting Factor

The summit will also address one of the less glamorous but more consequential components of enterprise AI: data.

The “Knowing Where to Bend and Where to Hold the Line: Data Readiness in the AI Era” session will bring together data leaders from VML, Sector Alarm Group, Carlsberg Group and Shell to discuss how organizations can balance imperfect data with governance and accountability.

That challenge is particularly relevant for AI agents. An employee may recognize that a report contains questionable information; an autonomous system may not.

Enterprises therefore need mechanisms for identifying trusted data, managing permissions, monitoring model behavior and establishing accountability when automated decisions produce unexpected results.

The event's focus on data readiness reflects a broader industry reality: AI models may be improving rapidly, but enterprise data environments remain fragmented across legacy systems, SaaS platforms and departmental databases.

The Autonomous Enterprise Needs New Operating Models

The second day will expand the discussion beyond AI implementation.

Diageo CTO Colin Shenoy will examine whether the move from digital transformation to AI transformation is genuinely different, while Reckitt's German Faraoni Heidenreich will discuss how AI is changing the economics of previously uneconomical business processes.

Other sessions will explore delivery discipline, AI partnerships, procurement and the impact of the EU AI Act on enterprise programs.

Those topics point to a broader transformation underway. Traditional technology procurement assumes that organizations buy relatively stable software with predictable functionality. AI systems are different: models evolve, capabilities change rapidly and the quality of results can depend heavily on data, prompts, context and integration.

That makes traditional RFP and procurement frameworks increasingly difficult to apply to AI deals.

The ISG Startup Challenge will provide another view of the emerging ecosystem, featuring startups working on agentic payments, AI trust and organizational systems for deploying AI safely at scale. Audience voting will determine which solution participants would be most likely to implement.

Market Landscape

Enterprise AI is shifting from experimentation toward operational deployment, but the market remains fragmented across foundation models, cloud infrastructure, AI agents, data platforms and enterprise applications.

For technology leaders, the competitive landscape increasingly includes hyperscalers such as Google, Microsoft and Amazon, alongside enterprise software companies such as Salesforce and Adobe. Their strategies differ, but the direction is similar: integrate AI more deeply into existing business workflows and data environments.

The rise of agentic AI could accelerate that convergence. Instead of simply generating text or analyzing information, enterprise agents can potentially coordinate tasks across CRM, ERP, marketing, finance and customer-service systems.

That creates a new category of enterprise infrastructure in which governance, identity, observability and data quality become as important as model performance.

Strategic Outlook

The 2026 ISG AI Impact Summit comes at a point when enterprises are confronting the difference between adopting AI and becoming an AI-enabled organization.

The companies likely to gain the most from autonomous systems will not necessarily be those deploying the largest number of AI tools. They may be organizations that redesign workflows, establish clear decision rights and build reliable data foundations before increasing machine autonomy.

For CIOs, CMOs, data leaders and enterprise architects, the emerging priority is therefore less about adding another AI application and more about determining where AI should act independently, where humans should remain responsible and how both sides should operate within a measurable governance framework.

Top Insights

• ISG's autonomous enterprise agenda signals a shift from AI experimentation toward redesigned workflows, accountability structures and measurable business outcomes.
• Sovereign AI is becoming an enterprise governance issue as organizations confront data ownership, security, compliance and dependence on external AI infrastructure.
• Agentic AI raises the importance of trusted enterprise data because autonomous systems can potentially execute workflows rather than simply generate recommendations.
• AI transformation is forcing companies to reconsider procurement, partnerships and operating models built for conventional software and digital transformation programs.
• Enterprise AI success increasingly depends on governance, data readiness and workflow redesign rather than model capabilities alone.

Get in touch with our MarTech Experts