AI is changing how quickly enterprises can make decisions across the customer journey, but faster decision-making is creating new questions about CX quality and accountability.
In fact, recent TheyDo research found that 68% of enterprise decision-makers believe AI has made decisions faster and more confident, yet 38% say AI-driven decisions have also created unintended CX problems.
To ensure they remain customer centric, CX leaders must assess AI by the quality and customer impact of the decisions it supports over speed or automation alone.
Jochem van der Veer, CEO at TheyDo, told CX Today that, whilst effective, AI does not remove the need for human accountability over its decisions.
“AI has compressed the time between the question and the answer," he said.
"But it hasn't removed the responsibility for what happens next.”
Speed Without the Full Picture
The growing use of AI in decision-making is creating a growing gap between confidence and quality, revealing that faster, more assured decisions do not always translate into better customer outcomes.
In fact, the Decision-Confidence Gap report revealed that 51% of businesses have seen complaints or support queries increase from AI-driven-decisions, with 42% experiencing a decline in customer trust.
These stats are reportedly driven by the fact that 77% of leaders say they now feel expected to make decisions more quickly because AI enables them to do so, creating a tendency to treat speed as evidence of progress.
Whilst AI can accelerate a decision without providing broader context on its potential impact across the customer journey, that context becomes particularly important when decisions are made within fragmented organizations.
“Faster and more confident doesn't mean better,” Veer highlighted.
“Most enterprises don't necessarily have an AI quality problem. The models are pretty good. But they have a decision-making problem.”
Teams using AI to optimize individual touchpoints may likely create friction in another area of the journey, meaning AI may identify a genuine change in a metric, but the response can still be wrong if teams lack visibility into the factors driving that change.
This reflects a wider concern that AI can produce an accurate answer based on the data available while still delivering a poor outcome if it lacks the broader customer context
Fragmentation can make the consequences of errors harder to contain as the issue increases across an organization.
“The risk isn't necessarily scaling AI," he explained.
"It's scaling AI while the organization underneath it remains fragmented.”
This means ensuring AI-supported decisions are made with enough customer and operational context to understand what is happening across the journey rather than reacting to individual signals.
Turning Signals Into Safeguards
The next step for CX teams is to make decision quality part of how they measure AI success, closing the gap between AI systems, frontline, and operational teams for a broader view of customer context.
In fact, only 42% of businesses say they can identify CX issues proactively before customers are affected, while roughly a third discover problems through contact-center complaints.
With complaints being a lagging indicator of experience quality, this reveals that friction has already occurred when customers report an issue, meaning the organization may already have damaged trust with additional work for its service teams.
To close this gap, organizations must bring together customer intelligence signals such as feedback and sentiment, as well as operational metrics, business priorities, and information about the end-to-end journey.
A recommendation based on one dataset may appear reasonable, but its implications can change when viewed alongside what happened earlier.
This requires teams to understand why an AI system reached a particular recommendation, becoming increasingly important as AI moves into decisions with direct consequences for customers.
Veer noted:
“Being intentional about making the recommendations traceable is another essential thing.”
AI traceability offers teams a way to inspect the reasoning and evidence behind recommendations to identify where a decision may have relied on incomplete context.
For service team managers, this offers clearer accountability for both the AI and employees when decisions cross organizational boundaries, as effective AI-supported decisions cannot always sit within a single function.
Furthermore, the report revealed that organizations with AI widely embedded are 3.5 times more likely to report full visibility of the customer journey than those using AI in isolated areas, suggesting that the measure of progress should be how effectively it connects decision-making.
“The next competitive advantage isn't necessarily access to AI, but it's the ability to make like consistently better decisions with it,” he concludes.
For CX leaders, the next practical priority is to measure AI by the quality and consequences of the decisions it supports, offering access to connected customer context, making recommendations traceable, and establishing ownership for decisions that span multiple teams.
From here, organizations can confidently identify and address customer friction before it becomes a complaint, creating a clearer link between AI deployment and measurable improvements.
Check out the full interview with Jochem above to find out more about how your organization can improve its decision-making.
