A new Salesforce survey has found that although 84% of CIOs believe Artificial Intelligence (AI) will be as revolutionary for businesses as the internet, only 11% have fully adopted the technology. The survey, which involved 150 CIOs from enterprises with 1,000 or more employees, sheds light on the key obstacles preventing broader AI implementation, including data infrastructure and security concerns.
AI knowledge gaps and a cautious approach to adoption
The survey reveals that many CIOs feel under pressure to be AI experts, with 61% stating that they are expected to know more about AI than they do. Much of their knowledge is gathered from peers and industry sources. While 84% of CIOs agree that AI will have a significant impact on business, 67% are taking a more conservative approach to implementing it compared to other technologies.
Current spending reflects this cautious attitude, with CIOs allocating a median of 20% of their budgets to data infrastructure and only 5% to AI. Concerns over security and privacy risks, along with the lack of reliable data, are the main reasons for this restraint. Many companies are prioritising data projects before increasing their AI investment.
“Generative AI is one of the most transformative technologies of this century,” said Juan Perez, CIO of Salesforce. “This research provides a glimpse into the foundations CIOs are laying across industries to prepare for AI.”
Mismatched expectations and “shadow AI” risks
Despite strong executive support, with 77% of CIOs reporting solid backing for AI's potential, many believe there is a gap between expectations and reality. Some 68% of CIOs feel that their business partners have unrealistic expectations regarding the timeline for return on investment (ROI) from AI initiatives. This urgency has led to a rise in “shadow AI”, where employees are using AI tools without official approval. Departments such as sales, marketing, and e-commerce report that they have “fully implemented” AI, but much of this activity occurs without oversight, creating serious security risks as sensitive information is sent through unsecured systems.
“We don't yet know how to use AI among the larger employee base and how to control its governance,” said a CIO from the professional services sector in the United States.
While certain departments like customer service are seen as having clear AI use cases, they are also viewed as being less prepared for AI adoption. In contrast, marketing teams are enthusiastic about AI but often lack the skills to implement it effectively.
Data security and infrastructure must be addressed first
Before moving forward with AI projects, many CIOs are tackling critical data infrastructure and security issues. Concerns over outdated systems, legacy applications, and the accessibility of data are slowing down the AI rollout. The security of AI tools is a top priority, and CIOs are looking for vendors that offer strong safety measures and ethical safeguards.
“The biggest barrier to AI implementation is data accessibility. Legacy systems and outdated infrastructure can hinder the effectiveness of AI,” said a CIO from the manufacturing sector in the United States.
“We have lots of old data centres and legacy applications that don't natively support data exchange,” said another CIO from the financial services sector in Italy.
CIOs are allocating increasing portions of their budgets to data-related initiatives, recognising that a strong data foundation is crucial before scaling AI efforts. However, there is still uncertainty about how much to invest in AI at this early stage, with only 47% of CIOs confident that they are allocating the right amount to these projects.
To bridge the gap, many CIOs are launching small-scale AI pilots to demonstrate the technology's value. Around 75% of respondents describe their organisations as being in the experimental stage of AI adoption. These pilot projects are seen as essential for showcasing AI's benefits and gaining broader organisational support.
As the AI landscape evolves, CIOs are relying on their peers for guidance. Despite seeking advice from analyst firms and vendors, CIOs say they most trust their fellow executives when it comes to AI insights.
“This is the perfect opportunity for knowledge sharing,” said Perez. “We are all navigating this AI revolution in real-time, and it's essential to exchange our learnings so we can build successful AI enterprises.”
