The vast majority of organizations do not have data that can be used for artificial intelligence (AI), according to research from big data provider Cloudera.
The results presented in the global report The Evolution of AI: The State of Enterprise AI and Data Architectureshow that only 9% of IT executives surveyed said that all their data is accessible and can be used for AI, while 38% reported that most of their data is accessible.
surveyThe study, commissioned by Cloudera and conducted by market research consultancy Researchscape, surveyed 1,574 IT executives from Europe, Middle East and Africa, the United States and Asia Pacific working in companies with more than 1,000 employees. The study was conducted in July 2025 and its results were published this week, coinciding with the London version of the provider's report. Evolution25 event.
At the same event, Cloudera also released a UK study exclusively for the conference, which found that a lack of diverse gender representation in AI leadership positions will lead to biased AI results.
A UK survey of 100 female IT decision makers by research firm Sapio found that more than two thirds (68%) are concerned about the under-representation of women in senior AI roles. More than half (56%) believe a lack of representation will lead to bias in AI results, with 57% saying AI is inherently biased because the CEOs of artificial intelligence companies are predominantly male.
Mary Wells, chief marketing officer at Cloudera, said: “Gender inequality remains a major challenge in the technology sector. Despite progress, we still face significant barriers.
“Artificial intelligence is a catalyst for positive change that can transform businesses, industries and economies. However, without the participation of diverse groups in AI design and strategy from the very beginning, we risk perpetuating old biases.”
Some 82% of respondents believe a gender imbalance still exists in the tech sector, with four in five (80%) believing gender equality is a superficial tick at their company.
Artificial intelligence is a catalyst for positive change with the potential to transform businesses, industries and economies. However, without the participation of diverse groups in AI design and strategy from the very beginning, we risk perpetuating old biases.
Mary Wells, Cloudera
The biggest challenges holding women IT leaders back from AI leadership roles were gender bias in hiring and promotions (68%), limited development opportunities (66%) and, according to the global study, data that is not AI-ready (60%).
Regarding global data readiness for AI research, Sergio Gago, CTO of Cloudera, said: “Over the past 12 months, AI has evolved from a strategic priority to an urgent concern, actively changing operations and redefining the rules of competition.
“However, our research shows that challenges around security, compliance and data use remain. Organizations need access to all their data, wherever it is and in any form, to manage it securely and gain predictive insights in real time.”
The majority of respondents (86%) said they are “data-driven.” But they also face significant challenges when it comes to supporting AI workloads.
Data integration was cited by 37% as the biggest challenge. Also problematic are storage performance (17%), computer power (17%), lack of automation (17%) and latency (12%). About 61% of respondents said that data silos have hampered their ability to scale their AI efforts.
Perhaps unsurprisingly protection of data used in artificial intelligence systems turned out to be a pain point for those surveyed. While 77% expressed confidence in their organization's ability to protect data used for AI, the study found significant concerns related to data leakage during model training (50%), unauthorized access to data (48%), insecure third-party AI tools (43%), lack of transparency or explainability of model output (39%), and manipulation of the model or its poisoning (35%).
However, according to the study, the rate of adoption of AI is high: 21% of respondents said that AI is already fully integrated into their business processes, 54% said that AI is significantly integrated, and 21% said that it is somewhat integrated.
About 70% of respondents said they had already made significant progress in implementing AI initiatives.
