F5 has presented the Conclusions of your report on the status of the AI 2025 Application Strategywhich reflects that only the 2% of organizations are prepared correctly to climb the AI safely in all its operations. The text measures six operational maturity factors, including the safety and alignment of infrastructure for this technology.
The report, made from an investigation that includes data from 650 responsible for IT and 150 of the strategy of the companies around the world with billing of more than 200 million dollars, shows the complicated situation of the preparation of companies for AI, as well as their ability to adapt to new trends.
Although 77% of companies show moderate preparation for AI, most do not have a solid or cloud security, which exposes them to risks. Meanwhile, 21% of companies have a low preparation, which limits its competitiveness.
70% of moderately prepared organizations use the generatively actively, and practically all others are working to use it. On average, 25% of applications use AI. As for highly prepared organizations, they often use AI in a much higher percentage, and it is expected to reach their entire portfolio.
Organizations with low preparation use AI in less than 25% of their applications. Generally, they do it in isolated or experimental environments. As for organizations with moderate preparation, they currently have AI in about a third of their applications.
65% of respondents use two or more payment models and at least one Open Source. On average, each organization uses three models, and the use of several is related to the implementation in more than one environment or location. Most of the models in use are currently paid, but the Open Source alternatives are also popular. The main models of this type cited in the survey are the Meta flame variants, the variants of the Mistral and Gemma of Google.
On the other hand, the report highlights cybersecurity concerns as companies expand their AI capabilities, and shows a worrying reality about the capacity of companies to address the complexity of the protection of AI workloads. But 71% of respondents already use AI to improve security, since companies consider this technology as a viable cybersecurity asset.
Only 18% of moderately prepared companies have implemented an AI Firewall, while 47% aspire to implement it in a year, which shows that specific protections are missing for AI. In addition, only 24% of organizations perform continuous data labeling, indicating lower transparency and a higher risk of adversary attacks.
Hybrid cloud environments create governance gaps due to inconsistencies between clouds. This exposes workflows and data to vulnerabilities. In addition, the use of various AI models increases risks without adequate control frames for Open Source tools, and have to face an enlarged attack surface.
For F5, the key actions for companies to improve the scalability and safety of AI include the diversification of AI models, and the expansion of the use of AI in all workflows, overcoming pilots and integrating it into operations, analytics and security to transform companies. In addition, it advises to integrate specific security for artificial intelligence, with protections such as AI firewalls, as well as the formalization of data governance processes, including its labeling.
Aran Erel, general director of F5 Spain, has pointed out that «As IA becomes the nucleus of the business strategy, preparation requires more than experimentation: it demands security, scalability and alignment. This report highlights the practical steps for organizations to implement the AI with confidence. IA is already transforming security operations, but without mature governance and specific protections, companies run the risk of increasing threats«.
