Gartner predicts surge in AI transparency investments by 2028

Growing enterprise adoption of generative AI is accelerating demand for explainability and observability tools, as organisations prioritise trust, accuracy, and governance to scale AI deployments responsibly across business-critical environments.

Global research firm Gartner has projected a sharp rise in investments aimed at improving transparency in artificial intelligence systems, particularly in generative AI (GenAI). The firm estimates that by 2028, nearly half of all GenAI deployments will incorporate explainable AI (XAI) and large language model (LLM) observability capabilities—up significantly from current levels.

Explainable AI refers to a framework that helps organisations better understand how AI models function, including their decision-making processes, strengths, limitations, and potential biases. This growing focus on transparency is being driven by the need to ensure fairness, accountability, and reliability in AI-led outcomes.

Rising need for trust in AI systems

As enterprises increasingly integrate GenAI into operations, the demand for trust is outpacing the technology’s rapid evolution. According to Gartner, XAI provides clarity on why a model generates a particular response, while observability tools assess how those outputs are produced and whether they are dependable.

LLM observability platforms go beyond traditional IT monitoring by evaluating critical AI-specific parameters such as hallucinations, bias, token usage, and response quality. These tools are now being used not only by developers but also by IT operations and site reliability engineering teams to maintain system performance in real-world environments.

Gartner warns that without strong explainability and observability frameworks, organisations may limit AI usage to low-risk or non-critical scenarios, ultimately constraining the technology’s return on investment.

Expanding market and governance priorities

The research firm forecasts the global GenAI market to surpass $25 billion by 2026 and grow to $75 billion by 2029, reflecting widespread adoption across industries. However, this expansion also brings heightened risks, including inaccuracies, biased outputs, and unreliable decision-making.

To address these challenges, organisations are increasingly shifting focus from basic performance metrics such as speed and cost to deeper quality indicators like factual accuracy and logical consistency. This evolution is also driving the adoption of governance-focused practices, including human validation of AI outputs.

Gartner recommends that enterprises implement structured approaches such as XAI tracing for critical use cases, comprehensive observability platforms, continuous model evaluation within development pipelines, and broader stakeholder awareness around AI governance requirements.

As AI systems move beyond experimental environments into large-scale deployment, Gartner emphasises that combining explainability with robust monitoring will be essential to unlocking the full business value of generative AI.