What CIOs Need to Know About Deploying AI

June 05, 2023

Contributor: John Hillery and Nathan Lewis

These five points are essential for making important leadership decisions.

Even before ChatGPT, one-third of CIOs say their organization had already deployed artificial intelligence (AI) technologies, and 15% more believe they will deploy AI within the next year, according to the 2023 Gartner CIO and Technology Executive Survey. But deciding how best to proceed means factoring AI into business value, risk, talent and investment priorities.

How to manage AI expectations: Five things CIOs need to know

Business leaders have high expectations about AI that CIOs will need to manage. CIOs need to be fluent in the technical language of AI as well as the risks and opportunities for their business.

Here are five things every CIO should know about the AI landscape to become a successful business leader in the rollout of AI.

No. 1: Prior to ChatGPT, most current deployments of AI focused on efficiency improvement scenarios at the business unit level, rather than enterprisewide.

Most organizations typically deploy AI in a business unit or area for the following use cases:

  • Smart process automation and robotics systems

  • Automating and personalizing at scale

  • Increase workforce productivity and AI-enabled decisions accuracy

No. 2: Best use cases for AI require access to a sufficient amount of reliable data that is relevant, logical and high quality.

CIOs often expect AI to add value to the business but must be clear on what’s feasible. Most AI business value is generated from one-off, point-to-point solutions. Getting more value from solutions at scale may require deep business process changes, and new ways of working between AI teams and software engineering, because AI is difficult to integrate into existing systems.

The following use cases are both highly feasible and highly likely to drive business value, so investment here will be easy to justify: 

  • Price optimization

  • Lead scoring

  • Demand generation

These are examples of highly feasible AI use cases for which the business value is likely medium, so investment will be more opportunistic:

  • Cross-selling and upselling

  • Territory organization

  • Sales content personalization

  • Knowledge management

  • Account intelligence

No. 3: The growing usage of generative AI opens organizations up to legal issues over copyrighted or protected content and confidential information security breaches.

Generative AI can augment and accelerate multiple business capabilities, but CIOs need to be aware of emerging government regulations and frameworks around AI, especially as increased usage triggers more questions about ethics and responsibility. The following risks are associated with AI and generative AI. 

AI risks:

  • Regulatory. AI poses legal risks by potentially opening up organizations to lawsuits over copyrighted or protected content, information and data.

  • Reputational. AI can amplify biases and create a “black box” — an AI system with no user visibility into inputs and operations.

  • Competency. AI requires a unique set of skills that need to be intentionally sourced through upskilling existing talent or from academia or startups.

  • False output. Generative AI, and ChatGPT specifically, can be unstable, be erroneous in reasoning, can fail to comprehend the entire context, has limited explainability and trackability, and is biased.

  • Security. Your sensitive data and intellectual property can be used to generate responses to users outside the organization — such as service provider employees and hackers.

  • Legal. Generative AI can present legal risks associated with intellectual property and privacy concerns, including copyright infringement, trade secret misappropriation, data privacy, model bias and model security.

No. 4: You do not need to build your AI. You can acquire AI and use embedded AI in existing applications.

There are many ways of acquiring AI outside of internal development, such as enterprise applications you are already using, packaged applications you can buy and AI add-ons (chatbots, virtual assistants, etc.). Organizations can:

  • Buy APIs (e.g., Amazon Web Services, Google, IBM, Microsoft) and packaged applications (e.g., IBM, Microsoft, Oracle, SAP, SAS)

  • Build open source (e.g., Python, Apache Spark, TensorFlor), data science/machine learning platforms, citizen data science tools

  • Outsource to global and/or local consultants, specialists and/or systems integrators

CIOs indicate that AI talent is not a major resource concern, and they combine both internal and external hiring to source talent needed for successful AI deployment. Four roles are key, though: data scientists, data engineers, AI engineers and business experts.

No. 5: Responsible AI adoption and journey planning means engaging with stakeholders to understand current efforts and defining what AI means for your corporate values.

Next steps for CIOs

  • Create a succinct AI strategy document that synthesizes your vision and potential benefits, audits and mitigates risks, captures KPIs, and outlines best practices for value creation.

  • Identify sponsors for AI projects and ensure their KPIs are being measured accurately and communicated widely.

  • Invest in data literacy programs to instill a data-driven culture.

  • Instill responsible AI practices and make them foundational to your AI strategy, not an afterthought.

John Hillery is Managing Vice President in Peer and Practitioner Research for the Gartner CIO Research Group. His current research focus is IT strategy, governance, operating models, performance measurement, and talent and evolution of the CIO role.

Nathan Lewis is a specialist in the Gartner CIO & Industries Group. His current focus is on peer and practitioner research.


Experience CIO and IT Executive conferences

Join your peers for the unveiling of the latest insights at Gartner conferences.

Drive stronger performance on your mission-critical priorities.