March 30, 2026
Could Rising Demand for XAI Improve Trust Issues in CX?
New research from Gartner predicts that, by 2028, the growing importance of explainable AI (XAI) will drive large language model (LLM) observability investments to 50% of GenAI deployments, up from 15% today. It is a forecast that reflects mounting pressure on enterprises to justify how their AI systems arrive at decisions. For a CX industry already grappling with a widening trust gap between consumers and AI-powered services, the timing could hardly be more relevant.
Gartner’s Findings
Gartner defines XAI as “a set of capabilities that describes a model, highlights its strengths and weaknesses, predicts its likely behaviour and identifies any potential biases”. Explainable AI can describe the functioning of a model to ensure it is transparent and accurate in its decision making.
LLM observability, on the other hand, is used to validate how that response was generated and whether it can be relied upon. Pankaj Prasad, Sr Principal Analyst at Gartner claims that traditional observability has focused primarily on speed and cost. The priority for modern LLM observability solutions is instead moving toward quality measures, including factual accuracy, logical correctness, and resistance to sycophancy, alongside governance-focused evaluation methods like human-in-the-loop content verification.
Essentially, therefore, the growing pressure to know why a model acts in the way it does via XAI could also lead to greater research into how AI is generating its content using LLM observability. Both XAI and LLM observability are critical, Prasad explains: “Explainability turns a GenAI output into a defensible, auditable insight. LLM observability ensures the model behaves as expected over time. Without both, GenAI cannot mature beyond controlled lab environments.”
The research firm forecasts rapid expansion in the underlying market. The global GenAI models market, it says, is expected to exceed $25 billion in 2026 and reach $75 billion by 2029. As adoption scales, so does demand for the mechanisms that protect against factual inaccuracies in generative AI.
To improve trustworthiness of enterprise generative AI solutions, Gartner recommends mandating XAI tracing for all high impact use cases, LLM observability across key metrics, and continuous LLM evaluation across integration and delivery pipelines. It also suggests keeping stakeholders informed about explainability requirements.
The CX Trust Problem
A study published on ResearchGate, specifically connects the need for explainable AI with the customer experience industry. Focusing specifically on AI-driven customer interactions, it found that a lack of explanation leads directly to user scepticism and disengagement. The paper identifies interpretability, transparency, and accountability as the core pillars that XAI must address if AI systems are to earn sustained consumer confidence in commercial settings.
Another peer-reviewed study published in The American Journal of Engineering and Technology, examining XAI implementation in banking, insurance call centres, and online retail, found that transparency and interpretability of model outputs positively influenced customer trust and loyalty while improving internal operational efficiency. The banking and retail deployments in particular produced measurable gains in retention, conversion, and satisfaction.
Although consumer trust in AI is low right now, losing businesses both loyalty and revenue. If you are able to demonstrate that your generative AI is indeed trustworthy, the upside spoils of this could be huge, with customers also willing to pay more for brands they can trust their data with. Organisations that invest early in explainability and observability will not only meet incoming governance expectations, but they may also find themselves on the right side of what customers are looking for.
