The pressure is on for CTOs to accelerate AI adoption and demonstrate meaningful returns. Despite this, the vast majority of organisations remain at an early stage of implementation, either laying the technical foundations or running pilots that have yet to scale. Only 12% of enterprises are considered “AI leaders” building architecture for enhanced, autonomous processes at enterprise scale.
The industry’s largest technology providers collectively spent $155 billion on AI in 2025 alone, while software subscription costs have risen by as much as 37% by the end of 2025, driven largely by AI-related pricing adjustments. Boards and C-suites are demanding returns on these ever-growing investments, yet more than 40% of agentic AI projects are set to be cancelled by the end of next year.
Implementing AI correctly is also incredibly important for businesses looking to gain an advantage and differentiate versus their competitors. However, successful AI investment needs buy-in across departments, with different requirements at each stage of delivery. How can CTOs build cross-functional alignment within the enterprise needed to scale AI and achieve meaningful ROI?
Securing the Stakeholders
Approval for AI investments and deployments are rarely confined to a single department and typically needs to pass scrutiny from finance, compliance, legal, and IT. AI investments will fail unless there is coordinated buy-in from every level, but each department will have specific requirements that need to be met.
92% of CFOs and finance teams report feeling pressure to demonstrate ROI on AI investments, but may feel hesitant to trial more AI after being previously burned by technology spend that failed to deliver, so vague promises of efficiency gains are unlikely to work.
Compliance and InfoSec teams will, rightfully so, approach AI investments with caution, which needs to be balanced with the organisation’s objectives. Data sovereignty and regulatory alignment are non-negotiable requirements and especially important if the organisation is global, as any AI deployments will need to comply with regional regulations, such as the EU AI Act, as well as data protection obligations such as GDPR.
CIOs and developer teams are thinking about architecture. AI vendor lock-in is a persistent concern that could result in future inflexibility, and any AI development that cannot integrate with the existing technology stack creates technical debt. 32% of IT leaders said integrating AI into their existing workflow was their top technology-related challenge.
Walking Before You Run
One size won’t fit all when it comes to overcoming obstacles to achieve buy-in, with challenges varying depending on sector requirements and the unique protocols of individual companies. However, a natural starting point for all companies should be short-term pilots. When designed well, they can create momentum for further developments. Pilots help to build on proof of concepts by deploying into a real environment, with real data, real customers, and real users, and reveal how AI can improve productivity and efficiency in certain tasks. Equally, they can reveal areas of improvement before scaling.
A well-run pilot also helps to generate the evidence that each stakeholder requires for their buy-in. After initial approval, buy-in from across departments keeps momentum going, with regular progress reviews. AI orchestration takes this further, and helps to continue to enhance what could be an efficient AI process by allowing the organisation to switch out AI providers or tools as new ones come onto the market, such as better language translation or accent detection.
Learning From Early AI Adopters
Customer experience (CX) is typically the first business department to explore AI solutions, which means it often leads the way for wider enterprise adoption. CX embraced generative AI earlier than most sectors, and is now looking into how more autonomous agentic workflows could improve productivity even further.
Having experienced generative AI, CX leaders will be better placed to present use cases for further AI funding. One of the key learnings was to ensure the return on investment and productivity gains were clear. Those that didn’t benchmark couldn’t provide a view of ‘before’ and ‘after’, making it harder to provide a positive use case.
CX is uniquely positioned within the business as it brings together the right data, processes, governance and people to enable efficient AI driven outcomes, and people to enable efficient AI, so should continue to be tracked to see how it is adopting cutting-edge solutions and how it is refining its approach to ensure success.
To Buy-In or not to Buy-In?
AI investments require buy-in from different departments across an organisation, including C-level oversight, with different requirements needed to reach each stage of buy-in. Each team and stakeholder needs to be engaged with on their own terms to understand what constitutes “success” for their department.
Businesses that invest in achieving alignment early in the process and ensuring AI orchestration as a broader consideration will be part of the 12% of AI leaders breaking out of pilots and into scalable AI maturity.
