Editorial
The Gist
- Rapid adoption. Generative AI in customer experience is quickly gaining traction, with one-third of organizations already using it in at least one business function.
- CIO concerns. Security, privacy and data governance are top concerns for CIOs as they strategize the integration of generative AI into customer experience.
- Uncharted territory. While generative AI offers abundant opportunities for transformation, it is largely seen as a tool, not a strategy, requiring careful planning and governance.
When ChatGPT debuted last year, it captured the imaginations of CMOs and those in charge of customer experience. Clearly, customer experience represents one of the most significant applications for large language models (LLMs). However, how fast will applications of generative AI in customer experience occur?
A recent McKinsey Global Survey says that “less than a year after many of these (LLM) tools debuted, one-third say their organizations are using generative AI regularly in at least one business function.” Nearly one-quarter of surveyed C-suite executives say they are using generative AI tools for work, and more than one-quarter of respondents from companies using AI say generative AI is already on the boards’ agenda.
Given this, where are CIOs in planning for generative AI in customer experience? How are they enabling their organizations especially around customer experience? What are their common concerns and perceived risks? And what is the CIO plan of attack for driving LLMs into their business’ transformation plan?
Is Generative AI in Customer Experience Part of Your Transformation Plan?
I wanted to start by understanding whether generative AI in customer experience has made the grade and then what do CIOs perceive as the greatest opportunities within their organizations? Clearly, business is booming at NVIDIA. Several CIOs said their answer is a qualified yes, but all access to these tools from company hardware not just the company network is restricted in many cases due to fears about personal identifiable information or IP leakage in prompt texts.
Manufacturing CIO Joanne Friedman started by saying, “There are abundant opportunities to leverage AI but the greatest ones — improving transformation ROI, revenue, and growth — are those where cost savings has already been achieved.” Meanwhile, Manhattanville College CIO Jim Russell says, “It is a yes on multiple fronts. It is content and tools we will use to expand our curricula and learning goals across most disciplines as well in the way we recruit, retain, and support our students as customers. So primarily what we will do is incorporate it in how we leverage AI tools.”
A little behind Russell are school districts and New Zealand CIOs. Chris Turner, a school district CIO says, “As a K12 organization, we're in the midst of examining how our peer districts are negotiating the challenges and opportunities of AI, and discussing its impacts on our local teaching and learning.” Meanwhile, Anthony McMahon says, “From what I've observed, it's not in the plan for most organizations over here. It's sitting on the radar and being talked about, but not actively planned for or how it will be delivered. The ones that are adopting it today are startups, seeing how they can use it to disrupt.”
Providing a cross industry perspective, Constellation Research VP Dion Hinchcliffe, says, “In my CIO conversations over the last few months, I’ve seen that Generative AI is in the plans of most IT organizations. It’s quite high priority, driven by boards who are worried about rapid disruption." The greatest opportunities are generally not the easiest ones. These include:
- Private LLM of organizational knowledge (hundreds of use cases, and pre-requisite for rest).
- Transformation of operations, R&D, strategy, customer experience, and supply chain.
Hinchcliffe continues by saying, “Marketing and customer service are clearly the low hanging fruit for generative AI. This is what many will tackle first with it, but they are not the greatest opportunities for most organizations.”
Related Article: AI in Business: How Company Leaders Are Taking the Plunge
Enabling Generative AI Experimentation
I asked next what will move organizations from experimentation into production regarding generative AI in customer experience? Russell said, “Like many, we lack muscles for agility and pivots, but will we work on this to survive. Some areas will move sooner into production like SEO, programming, and perhaps marketing or vendor supported initiatives including chatbots. Other areas lack the data governance to leverage their data. That will keep, in the world of public or vendor data, until we mature or incorporate the changing lessons of LLM into our own data governance program.”
For Turner, he said, “I would define our approach to this groundbreaking innovation as firmly cautious, and we are disabling and filtering until we land on board-approved policy guidance.” Meanwhile, Friedman suggests that CIOs should “enable experimentation with LLM to (chain of thought, prompt engineering) to ascertain the best use case which should go into production.” McMahon adds, “My advice is to remove LLMs from this entirely. The question should be how are you enabling your organization to experiment. This should apply to all technology. To be clear, LLMs are a tool, not a strategy. For this reason, experimentation is part of any strategy approach.”
Hinchcliffe agrees and suggests, “Shadow AI is real, and caused by the same forces that have unleashed Shadow IT. IT must get out in front of generative AI quickly. With a clear understanding and plan for not being able to do it all. Let’s be clear, most non-IT stakeholders severely underestimate the effort needed to deliver generative AI properly, addressing security, ethics, privacy, IP protection, responsible usage, bias and cost effectiveness.”
Related Article: 5 CX KPIs Companies Are Improving With AI
Risks Most Concerning CIOs
I wanted to know whether it is open source? Security? Privacy? At this point a healthcare executive said, “Security and privacy are at the top, but in a healthcare setting, any AI hallucination has serious potential to violate the first, do no harm doctrine.” For Russell, “It is very much equal between intellectual property (organizational or personal) and accuracy. Examples include the student-generated content that vendors house or process. And educational needs to have almost pristine data to preempt trust erosion.”
Meanwhile, Friedman says, “I would say, it is the date on which the model stopped being trained/learning and before IP leakage.” McMahon agrees by saying, “reliable, truthful outputs. People need to treat all outputs as an initial draft which needs to be validated and potentially edited. They certainly should not be spewing copy straight out to their client base.”
Summarizing, an industry perspective, Hinchcliffe suggests the biggest risks are:
- Ownership and control.
- Privacy and safety.
- Making sure IP can’t escape.
- Compliance.
- Factual accuracy.
- Vendor lock-in.
- Skills to maintain.
- Cybersecurity.
- Model “poisoning.”
Related Article: How to Pick the Right Flavor of Generative AI
Developing a Generative AI Plan of Attack?
Given the above, I wanted to know how CIOs should proceed with generative AI and LLMs. Most of generative AI in customer experience is new and additive to the data stack. Russell says, “define, align, fund, and execute. Define the scope of which areas can leverage or absorb LLM. You should ensure that it is aligned with strategic or tactical goals. Put it in the sieve for funding. Run with it.”
Friedman adds it is critical to do the following:
- Educate stakeholders.
- Ask why? What value will be captured or created and how it will be measured?
- Build risk profiles and analyze each use case.
- Align with lifecycle stage.
Webinar
Sep
25
CMS Briefing: A Live Look at What’s Next in AI-Driven Platforms
Learn how leading organizations are using AI‑driven tools to publish faster, personalize smarter and stay secure.
Register
Webinar
On demand
Ready or Not: How Data-First Organizations Are Unlocking Agentforce Potential
Learn how to cut through the noise, activate Agentforce and build a Salesforce AI strategy that actually delivers.
Watch Now
Webinar
On demand
AI in Customer Service: Faster Resolutions, Happier Customers
Don’t let rising demand burn out your team. See how to build a smarter, more resilient support org.
Watch Now
Webinar
On demand
From Hype to High-Impact CX Strategies That Actually Scale
Turn buzzworthy AI and outsourcing trends into measurable CX wins with fresh 2025 data.
Watch Now
Webinar
On demand
Insights to Action Rethinking the Contact Center for Real Business Impact
Join our exclusive webinar to hear CX executives share their innovative strategies for transforming service delivery.
Watch Now
Webinar
On demand
Accelerating Healthcare Ops with AI
Sign up to see how your peers are putting AI to work — and making it stick.
Watch Now
McMahon agrees and stresses that CIOs should determine "what's the problem they're trying to solve. LLMs are not an outcome, they are simply another tool which allows an outcome to occur.” With this said, Hinchcliffe says, “Unfortunately, governing LLM data is a bit trickier. Governance is an LLM risk. Steps to develop LLM capability should include:
- AI policy.
- Generative AI strategy.
- Data strategy.
- LLM platform definition.
- Vendor survey.
- Opportunity scanning.
- Proof of Concept, high value quick wins.
- Build/buy flexibility.
- Enterprise-wide ModelOps.
Parting Words on Generative AI in Customer Experience
Generative AI in customer experience is becoming part of many organization’s plans. It will have substantial impact on customer experience initiatives.
To succeed, plans need to have concrete business objectives and governance as core elements of their implementation. Governance should not be an afterthought. By doing this right, business will achieve their objectives without creating unintended consequences such PII or IP leakage.
Learn how you can join our contributor community.
