IBM Think 2026: AI, Governance and ROI

Automation has been introduced as a concept designed to almost be synonymous with efficiency. 

It leaves tasks that can be automated down to the machines; freeing up valuable human time, energy and resources.

At IBM’s Think 2026 event in London, the company revealed that in the past three years up until 2025, it saved US$4.5bn in productivity via the use of automation. 

Speaking at IBM Think 2026, Ana Paula Assis, Senior Vice President and Chair for IBM EMEA and APAC, says: “We did that not just for a cost-takeout objective, but also to make sure that we were reinvesting those productivity gains in innovation and growth for the company.”

With recent conversations around the value in AI from the finance point of view, IBM’s latest research exposes how ready corporations are with their AI use, and more importantly, the governance behind it. 

Ana further notes: “Starting with ROI, we are really starting to see some very credible use cases and situations where the benefits of using AI in organisational operations are clear”.

Financial risk and ROI

Gaps in AI governance carry financial risk, however IBM research reveals that only 37% have a formal and board-reviewed quantification of the risk. 

Craig Aston, CEO of Celerity, also speaking at Think 2026, says: “For internal use cases, projects must demonstrate the benefits and ROI, and they must be approved by the Governance Board before moving forward”.

New research from SAP notes that the average UK business is now spending £20.6m (US$27.1m) on AI, with returns set to rise by 39% by 2028.

Governance of AI 

Workiva’s recent survey in August earlier this year, highlights that AI errors have reached the boardroom. Internal audits caught errors that reached external audiences or board members at 26% of organisations.

IBM’s own research is displaying a higher proportion.

In a survey of 3,200 executives across EMEA, spanning eight countries, 80% of those surveyed admit to an AI-related incident reaching the C-suite in the last 12 months.

Already, IBM has found that across EMEA, a third of organisations (35%) are deploying agentic AI. 

However, despite its use, only 37% of respondents indicated that their current governance system is designed with agentic AI deployment in mind. 

This highlights the gap between governance and AI that is emerging in the C-suite. 

IBM’s ‘The New Realities of AI Governance’ report highlights that when it comes to accountability for adverse AI incidents, most C-suite leaders are expected to take responsibility. 

Out of the 80% who had an AI-related incident reach their boardrooms, 43% say it was formally reviewed at that level, whereas 37% say it was communicated informally.

So who should be responsible for AI at C-suite level? 

Assigning AI risk to the board doesn’t ensure complete accountability compared to a single person. 

IBM’s research notes that respondents voted no more than 17% for a single role, while 14.5% say it should be a shared equally among the C-suite.

• 17% think the CIO or CTO should be responsible 
• 14% think the CRO or CCO should be responsible 
• 13% think responsibility should lie with the Chief AI Officer 
• 10% think the AI vendor should be responsible
• 9% think it should be the CEO.

While nearly all (98%) of executives surveyed have formally assigned responsibility for AI risk at board level, only 26% have established a board-level AI risk committee. 

Leon Butler, UK CEO of IBM, speaking at Think 2026 notes: “The reality across AI deployment is that governance becomes less about contractual coverage and more about embedding governance and security directly into processes. 

“We are going to have agents communicating across different vendors, applications, and companies, so governance must be in place. As you deploy AI step-by-step, governance goes hand-in-hand with it.”

IBM Partners:

Amazon Web Services (AWS): Through a strategic two-way alliance, AWS and IBM deliver the framework, advanced software platforms and specialised technical expertise necessary to scale autonomous agentic AI across enterprise environments.

Microsoft: The IBM and Microsoft partnership focuses on delivering integrated solutions across cloud, cybersecurity, and AI. Combining Microsoft Azure with IBM’s Consulting services and Red Hat OpenShift, the partnership enables enterprises to modernise cloud infrastructures, enhance workforce productivity and streamline business processes through joint AI and enterprise software integrations.

SAP: Leveraging extensive consulting experience alongside SAP, IBM collaborates to unlock enterprise value across every stage of AI adoption. Together, the tech leaders design strategic roadmaps, co-develop innovative tools and drive RISE-supported business transformations, while deploying collaborative AI agents to power intelligent enterprise workflows.

Adobe: IBM has partnered with Adobe for over two decades. Currently, the partnership allows for an environment to innovate joint solutions with their products. Additionally, they collaborate on RedHat’s OpenShift, IBM watsonx and Adobe Experience Platform and CX Enterprise Coworker to deliver end-to-end agentic AI stack for regulated and complex environments.