Editorial
A Salesforce survey of 150 CIOs from companies with 1,000 or more employees finds there is considerable pressure on CIOs to become AI experts. Sixty-one percent of CIOs report feeling that they’re expected to have a greater understanding of AI than they possess, with many turning to their peers for information and insights they’re lacking.
Critically, 84% of enterprise CIOs view AI as a transformative technology, equating its significance to the rise of the internet. Despite recognizing its potential, 67% of CIOs are approaching AI with caution, signaling they are more measured in its adoption compared to other technologies. Furthermore, most CIOs believe their line-of-business stakeholders have unrealistic expectations about the speed AI will deliver a return on investment. This highlights the disconnect between the promise of AI and the pressures CIOs face in effectively managing its rollout and impact.
CIOs are under intense pressure to deliver faster results with generative AI, according to Salesforce CIO Juan Perez. Much of this is driven by a lack of “AI-savvy” business leaders. Perez emphasizes that clear business cases are essential, and generative AI initiatives should be treated like any other project. CIOs must also ensure their data infrastructure is in order, or they risk repeating the challenges faced during cloud adoption, where many found themselves playing catch-up. Success with AI hinges on starting with well-defined outcomes to guide the strategy forward.
AI Adoption
Only 11% of CIOs report having fully implemented AI, citing technical and organizational hurdles, such as security concerns and underdeveloped data infrastructure. Despite their broad technical expertise, CIOs are cautious, echoing insights from former BusinessWeek CIO Isaac Sacolick, who observes a significant gap between pilot projects and production deployments of AI. Sacolick estimates that for every concrete AI deployment, there are likely 10 exploratory projects, reflecting the early stages of AI integration across enterprises.
This cautious approach is evident in budget allocations, with CIOs prioritizing data infrastructure first. On average, the Salesforce research finds CIOs spend 20% of their IT budgets on data management and infrastructure, while AI rakes in just 5%. This emphasis underscores the importance of building a strong data foundation before leaning into AI. To be fair, there is a mismatch in AI readiness across departments. While customer service is seen as having the most potential for AI use cases, it is also perceived as the department least prepared to effectively handle the technology.
Applying AI
Importantly, the Salesforce research reveals that 77% of CIOs feel they have solid executive buy-in regarding the value of AI. However, despite this support, CIOs are concerned their business partners have unrealistic expectations. In fact, 68% of CIOs believe there is undue urgency in seeing a return on investment from AI, leading to a disconnect between business leaders' ambitions and the realities of AI implementation.
A significant number of leaders in several departments, like sales, marketing, service and e-commerce, claim they have fully implemented AI into their workflows. Yet, CIOs see a more nuanced picture. While customer service is perceived to have the most potential AI use cases, it is also viewed as the least enthusiastic about adopting the technology. In contrast, marketing teams are eager to leverage AI but are considered the least prepared in terms of skill sets and overall readiness. This mismatch highlights the complexity of integrating AI across different parts of an organization.
Perceived AI Risks
The slow pace of implementing enterprise-wide AI strategies is largely due to the preparatory work CIOs believe they must prioritize first. While there is enthusiasm for AI's potential, CIOs face significant hurdles, with security concerns and data quality issues topping the list. Security or privacy threats and a lack of trusted data rank as their biggest fears. Additionally, unsanctioned AI adoption is introducing new security risks, as employees may inadvertently expose sensitive information through unsecured language models.
Recognizing the need for a solid data foundation before fully embracing AI, CIOs are allocating, on average, four times more budget toward data initiatives than AI itself. Despite this investment, only 47% are confident they've allocated the correct amount to AI projects, reflecting uncertainty in the early days of the technology. Many CIOs struggle to define where and how AI should be integrated within their organizations, a challenge exacerbated by the novelty of AI and varying levels of understanding, or fear, across departments. As in-transition CIO Martin Davis points out, "ensuring data quality and trustworthiness is critical. Knowing the provenance of the data — where it came from, its validity and its source reliability — is crucial. Using unverified or poor-quality data can lead to flawed AI outcomes, encapsulated in the adage: AI + bad data = bad AI."
AI Pressure On CIOs
CIOs are nevertheless under immense pressure to define and execute AI strategies, as the technology's rapid evolution coincides with growing business expectations. AI presents not only a technical challenge, but also a strategic one, forcing CIOs to navigate unfamiliar territory with unprecedented pace. While 61% of CIOs believe their AI expertise is overestimated by stakeholders, only 9% feel their peers are more knowledgeable. This AI knowledge deficit spans across organizations, highlighting the shared learning curve CIOs face. Clearly, as Dave De Cremer indicates in “The AI Savvy Leader,” CIOs like other business leaders, “need to be just AI-savvy enough to recognize the benefits of AI for the organization and its stakeholders.” CIOs should have a cursory understanding of the workings of the technology too.
To close this gap, CIOs are turning to familiar resources, such as analyst firms, technology vendors and media outlets. However, they place the most trust in their peers from other companies, reflecting a collective effort to grasp the complexities of AI. “Although nearly 100% of CIOs are aware of AI and generative AI, only about half possess the deep expertise in data engineering, management, governance and security required to successfully implement these technologies,” according to Dion Hinchcliffe, VP of the CIO practice at The Futurum Group.
As Jim Russell, CIO of Manhattanville College notes, CIOs are accustomed to managing ambiguity, often relying on decades of experience to fill knowledge gaps. However, the speed of AI's advancement makes these gaps more critical, particularly when it comes to selecting the right models, large language models (LLMs) or small language models (SLMs), and aligning them with business needs. The challenge lies in determining which technologies will endure and how best to position their organizations amidst the ongoing AI revolution.
Salesforce offers three key recommendations for CIOs leading the generative AI charge: 1) establish clear guidelines for AI use to ensure responsible and effective deployment, 2) get your data house in order by industrializing your data infrastructure for seamless AI integration and 3) set clear metrics to measure success and guide decision making throughout the process. Without these, CIOs risk falling behind, much like the struggles experienced with cloud adoption.
In Conclusion
While AI holds immense potential for transforming businesses, CIOs are navigating uncharted waters as they balance strategic ambitions with the complexities of AI implementation. Despite widespread executive support and optimism about AI’s future, the road ahead is fraught with challenges — particularly in terms of security, data quality and realistic ROI expectations. CIOs are taking a cautious, data-first approach, recognizing that building a strong foundation is essential for the long-term success of AI initiatives. As they continue to learn and adapt, the roles of peer collaboration and knowledge sharing will be critical in overcoming these hurdles and shaping the future of AI in the enterprise.
Learn how you can join our contributor community.
