Why CTO confidence in scaling AI is declining

Chief technology officers are becoming less confident in their organizations' ability to successfully scale artificial intelligence. According to a recent Akkodis report, the share of CTOs confident in their companies' AI readiness has dropped from 82% in 2024 to 62% in 2025 and just 48% in 2026. Experts say businesses have moved beyond experimenting with AI and are now facing the far more complex challenge of integrating the technology into enterprise environments.

The main obstacle is no longer access to advanced AI models but their effective deployment across complex IT infrastructures. Organizations must integrate AI with legacy systems, fragmented data sources, existing risk management frameworks, and established business processes. It is during the scaling phase that companies discover whether they are truly prepared for enterprise-wide AI adoption.

While many organizations have successfully launched pilot projects demonstrating AI's potential, relatively few have been able to expand these initiatives into full-scale production. Meaningful AI deployment requires more than advanced technology- it also demands organizational change, workforce training, clear governance, and employee trust. Without these elements, even successful pilot programs struggle to deliver measurable business value.

The emergence of agentic AI has added another layer of complexity. Unlike traditional AI tools, agentic systems can write code, modify databases, update workflows, and automate business operations with minimal human intervention. This increased autonomy requires clearly defined responsibilities, human oversight, and robust governance mechanisms. Most organizations have yet to establish these operating models, limiting their ability to scale AI safely and effectively.

The report also highlights significant readiness gaps. Only 44% of CTOs believe their executive leadership has a sufficient understanding of AI, while just 36% are satisfied with employees' level of trust in the technology. The biggest barriers to AI adoption include a shortage of in-house technical talent (32%), uncertainty about return on investment (31%), and a lack of business urgency (27%).

Poor data quality remains another major challenge. Many enterprises continue to rely on fragmented systems containing outdated, duplicated, or incomplete information. Since AI performance depends heavily on the quality of its training and operational data, these issues often lead to inaccurate decisions. Once AI is deployed at scale, such errors can spread rapidly across the organization, creating significant business risks.

Despite these challenges, the report points to a major shift in how companies view digital transformation. Whereas organizations previously focused on reducing costs and improving operational efficiency, innovation has now become the primary driver of digital investment. Businesses increasingly see AI as a platform for creating new products, services, business models, and growth opportunities rather than simply optimizing existing processes. According to industry experts, organizations that successfully use AI to drive innovation-not just efficiency-will be better positioned to gain a lasting competitive advantage.