We know what artificial intelligence (AI) can do. But what should it be allowed to do and who should decide?
In 2027, these are just some of the questions that will force practitioners to maintain a balancing act between satisfying the needs of their business and their customers – who, incidentally, now have their own AI tools to deliver speed, convenience and easier journeys.
In the second of our three-part 2027 outlook series, CX Network looks in detail at six challenge AI could create for CX next year – and what CX and service leaders can do about them. Whether you run a contact center, an insights team, a VoC program, UX, or journey management, these insights can help you decide how to proceed in an increasingly complex landscape.
This article covers:
- AI can is not the same as AI should
- To lead this conversation, CX needs a seat at the table
- Beware of the AI slop – especially when it improves
- When you do use AI, don't let it take the praise
- Some customers now have their own AI to deflect organizations
- But not every customer loves AI
1. "AI can" is not the same as "AI should"
For smart ring creator Oura, AI is one of the most valuable tools in the kit – the company even developed its own metric to provide granular-level AI observability of the member experience.
Then there's the Oura Advisor, which transforms the ring's raw health data into actionable insights wearers can use to visualize their long-term sleep, activity, readiness and resilience data. In April 2026, this was followed with Oura's first-ever proprietary LLM within the Advisor experience, designed specifically for women's health.
One of the company's core values, however, is "human first" and, according to VP of member experience John Moses, this means understanding the journey moments where only humans can add value.
Moses explains: "The challenge for us and for lots of companies today is not simply can AI handle this, which it increasingly can, but should it handle this? Is it the right thing to do?
"Humans should stay in the loop at all times, not just to handle difficult cases, not just to handle escalations, but to continuously evaluate whether the AI itself is meeting the company standards, the values, the ideals," he adds.
2. To lead this conversation, CX needs a seat at the table
One way to ensure AI is handling the right things is for CX to be part of the decision-making committee.
Sue Duris, principal at M4 Communications, says: "Most enterprise AI decisions today are driven by technology and transformation teams, framed around capability and efficiency.
"The customer voice – what AI does to trust, to the people an automated path leaves behind, to the experience itself – is too often missing from the room where AI actually gets decided. And many CX leaders are quietly unsure where they fit, or whether they belong in that room at all," she continues.
Therefore, Duris says, the real challenge in 2027 will not come from the technology but from a lack of influence from CX on how it is applied. To shape AI decisions, practitioners need customer data, journey insights and the ability to answer the hard questions around trust and impact.
"That means being in the room when use cases are prioritized, when success is defined, and when the line between what AI handles and what a human must own gets drawn – not brought in afterward to smooth over what was already decided. It means forging the relationships with tech, compliance, product and operations now, before the decisions are made, not after," she adds.
"The organizations that get this right will be the ones where the customer had a seat at the table while it was being built," Duris concludes.
3. Beware of the AI slop – especially when it improves
Another danger lurks in giving AI too wide a mandate and Bill Staikos, founder and managing partner of Be Customer Led, calls it "synthetic sameness".
"We are already entering a copy-and-paste AI culture where people take LLM output and paste it directly to social media posts without a shred of original thought," Staikos says. "Similarly, companies will use the same models to write their content, answer customer questions, personalize messages, summarize research, design journeys and increasingly make decisions," he adds.
Today, this AI slop as it's known is easy to spot. But as AI improves, so too will the slop – and that isn't as positive as it sounds.
"Brands will sound the same, bots will behave the same, personalization will feel the same and experiences optimized by similar models against similar data will gradually converge toward the same answer," Staikos says.
"The CX challenge will become preserving distinctiveness, judgment, and humanity in a market where AI makes competence cheap and sameness incredibly easy."
4. When you do use AI, don't let it take the praise
CX's credibility challenge is nothing new. While many see the recent tech revolution as an opportunity for CX to shine, others understand that some corporate cultures can easily lead to an erosion of recognition for CX.
Jeannie Walters, CEO of Experience Investigators and author of Experience is Everything, says this means that in 2027, CX leaders will have to prove their value more clearly than ever.
"AI will be credited with cost savings and faster outcomes, so leaders will need to clearly articulate how faster insights led to better outcomes for customers and the organization based on their strategic direction and accountability," she says. "CX leaders have to be ready to lead as business strategists, not just as someone who reports on scores," Walters adds.
One way to achieve this is to draw a clear line to illustrate how CX drives the rest of the business.
Moses says: "Customer conversations contain signals about product, policy, operations, tooling, and CX teams need a way to surface those insights quickly enough for the broader organization to act on them."
5. Some customers now use their own AI to deflect organizations
In some organizations, AI has been used to acheive short-term KPIs, such as a reduction in average handle time, or a higher first-contact resolution (FCR). These quick wins have seen customers deflected and moments that should be human-led instead delegated to technology.
This year, the tables turned.
CX Network's 2026 research into the state of CX found the customers' use of AI for sales and service interactions was the third most influential CX trend and the third most important customer behavior for practitioners.
Mrunal Gangrade, VP of software engineering for JPMorgan Chase & Co says one 2027 challenge will be connecting with customers who are increasingly influenced and empowered by AI, and, as machine customers edge closer to reality, maintaining those connections.
"This creates a very different CX environment," Gangrade says of customer AI use.
"Businesses will no longer be designing experiences only for people interacting directly with their brand. Their AI systems may be interacting with customers who have already formed expectations based on AI recommendations, or eventually with another AI representing the customer."
On how to preserve those connections, Gangrade says customers still want to feel understood and respected and that businesses "will need to think carefully about trust, transparency, privacy, empathy, and customer choice as AI becomes more involved in these interactions".
She adds: "For me, the future of CX is not about choosing between humans and AI. It is about using AI intelligently while continuing to understand the person behind every interaction. The companies that can maintain that human connection while adapting to AI on both sides of the customer relationship will be better prepared for what comes next."
6. But not every customer loves AI
CX Network's 2026 research into the state of CX also found that awareness of how AI works and uses customer data was the leading customer behavior influencing the work of practitioners and for the first time, consumer privacy entered the top 10 CX trends.
Claire Cunningham, founder and MD of The Customer Connextion says this means CX has new competing interests to balance.
"Consumers continually want things to be faster, easier, and more personalized and companies want to continue to find ways to save money and increase revenue," she says.
"AI feels like the natural solution to that demand since human capital and training feels like a riskier investment over AI tools. However, consumers are increasingly worried about what is happening with their data."
Data isn't the only customer concern. Cunningham cites the "huge social media backlash" regarding data centers, the existential threat to humanity and the lack of guardrails that exist "to protect the very consumers that we're trying to meet the demands of".
Whether rooted in reality or not, Cunningham says those consumer concerns – and the general consumer perception of AI – are "causing a rapid erosion in customer trust for many brands".
According to Cunningham, trust is quickly becoming a primary indicator of brand perception and increasingly, revenue. "CX practitioners are going to have to work over time to prove the ROI of customer trust and discover what's contributing to its decline, namely the right balance of AI, security and human service.
"I think it's only going to continue to be an uphill battle for the CX'ers in 2027! But we can do it. If it's done right, CX and AI naturally work hand in hand but a lot of pieces, people, and regulations need to come together to do so," she adds.
Quick links
- 9 ways CX will change in 2027
- What does the future of customer insights look like in an AI-driven world?
- Responsible AI vs. AI governance: What CX professionals need to know