With the rapid adoption of artificial intelligence (AI) in various industries, Spanish companies highlight cybersecurity (33%) and the lack of data quality to train AI (32%) as their main concerns when implementing projects in this area . However, few companies are taking steps to improve data quality, limiting the success of their AI initiatives.
According to Hitachi Vantara’s “The State of Data Infrastructure Sustainability” report, data infrastructure and management are crucial to ensuring the quality needed to drive AI projects with positive business outcomes.
Key factors for AI success
In this context, information technology (IT) managers in companies identify two key factors for the success of AI projects: the use of high-quality data (35%) and adequate project management (39% ). These results underscore the importance of having well-managed, quality data, as well as trained professionals leading these initiatives.
Despite the recognized importance of data quality in the success of AI projects, the report reveals that the necessary data is available when and where required only 30% of the time. Furthermore, AI models developed by companies achieve accuracy in only 32% of cases. These results highlight the critical challenges organizations face in optimizing the use of AI and maximizing their business impact.
Data quality challenges
In a context where a third of companies surveyed express concern about the quality of their data, it is surprising that few organizations are taking concrete steps to address this challenge. Only 28% are actively working to improve data quality through accurate training of AI models.
Furthermore, 23% do not review the quality of the data used, and a significant 39% do not label them properly, a fundamental practice to improve data governance, as it facilitates its organization, traceability and accessibility, optimizing its value in the strategic decision making.
Return on investment and sustainability
Regarding the review of AI strategies, the study shows that many companies lack an analysis of parameters as relevant as return on investment (ROI) or sustainability. 65% do not consider sustainability as a priority when implementing their AI plans. Additionally, 63% of companies do not prioritize ROI when implementing solutions.
According to the most recent report of Hitachi Vantaraonly 28% of organizations are taking active steps to improve data quality, limiting the positive impact of AI on their business results
With sustainability as a key challenge in the digital era, it is striking that in Spain 86% of organizations are focused on the development of LLMs (larger general AI models), instead of opting for smaller specialized models. These large scale models have been shown to consume up to 100 times more energy. This large model approach is significantly higher than the European average, where it represents 64%.
Security and speed, key aspects
On the other hand, the technology managers surveyed highlight how AI implementation plans are carried out in companies, pointing out that almost half of the organizations prioritize two key aspects such as security (46%) and speed (45% ) in their projects.
In this context, security has become a priority due to the associated risks. 75% recognize that a significant data loss could be catastrophic for their operations. Additionally, 79% of respondents express concern about the use of AI to provide advanced tools to hackers.
“AI adoption largely depends on users’ trust in the system and the results. If your first experiences are clouded, your future capabilities are clouded,” says Antonio Espuela, Director Technical Sales EMEA West.
The role of data infrastructure in AI success
Despite recognizing data quality as one of the top factors for AI success (33%), many organizations lack the infrastructure to support consistent data quality standards.
A major challenge is that 79% of companies are testing and adjusting their AI solutions in real time without using controlled environments, increasing the risks of vulnerabilities and potential security flaws. Only 7% claim to use sandboxes for AI experimentation, raising concerns about the potential for security breaches and flawed results due to unreliable data.
Modern infrastructures can offer a solution to this problem, as they are not only more energy efficient, but also improve performance while reducing the carbon footprint. By adopting advanced, sustainable infrastructure, companies can improve data quality, reduce risks, and move toward environmentally responsible AI growth.
The value of having a trusted partner
As organizations advance AI initiatives, all IT leaders surveyed recognize that third-party support is essential to address critical areas such as:
- Hardware: Hardware effectiveness depends on its security, 24/7 availability, and efficiency in meeting sustainability goals. 20% of IT leaders identify a need for help developing scalable, future-proof hardware solutions.
- Data storage and processing solutions: Effective data solutions bring data closer to users and prioritize security and sustainability. The survey revealed that 29% of IT leaders need help with storing ROT (redundant, obsolete or trivial) data.
- Software: Secure and resilient software is key to protecting against cyberattacks and ensuring continuous access to data. 27% of IT leaders require external expertise to develop AI models and data virtualization solutions.
- Qualified personnel: Lack of qualifications remains a major obstacle, with 60% of leaders developing their AI skills through experimentation and 38% relying on self-learning, underscoring the need for specialized training and external support.
