Organisations are set to allocate more of their 2025 tech budgets to Generative AI (GenAI) than to security, according to new global research commissioned by Amazon Web Services (AWS). The Gen AI Adoption Index study, which surveyed 3,739 senior IT decision-makers across nine countries, was conducted by Access Partnership.
The survey’s findings show that 45% of organisations identified GenAI as their top IT investment priority for 2025. In comparison, 30% of organisations prioritised cybersecurity. Other areas, such as compute (13%), storage (7%), and physical hardware (4%), were lower in priority.
When assessing GenAI tools, 64% of respondents cited ease of integration into existing workflows as the most valued feature. This was followed by advanced capabilities at 48% and responsible AI guardrails at 43%.
Evolving AI leadership and deployment strategies
As organisations scale up their use of GenAI, leadership roles are also evolving. Although CTOs and Chief Innovation Officers continue to lead broader IT transformation efforts, AI-focused roles are gaining ground. In the AWS study, around 60% of respondents reported appointing a dedicated AI executive, such as a Chief AI Officer (CAIO). The figure is expected to reach 86% in 2026.
The study further stated that nine in ten organisations surveyed have already started deploying GenAI tools, signalling a broad shift from experimentation to integration. Although most organisations have started deploying such tools, adoption maturity varies. According to the study, 22% are in the exploratory phase, 25% are testing proof of concept, 21% are transitioning to production, and 23% have begun integrating GenAI into operational workflows. Only 8% reported having no adoption plans.
To meet the demand for AI skills, 56% of the firms are upskilling existing staff. An additional 19% of the organisations are expected to develop a GenAI training plan by the end of this year. Furthermore, 92% of respondents indicated plans to recruit for open job roles with GenAI skills.
As the organisations transition their GenAI workloads into production, they are increasingly integrating proprietary data with customisation capabilities. The research indicates that while 40% of organisations intend to use AI models directly off-the-shelf, 58% plan to develop custom applications using pre-existing models. Additionally, 55% aim to build applications on fine-tuned models leveraging proprietary data.
The survey was conducted in nine countries- the US, the UK, Canada, France, Germany, India, Japan, South Korea, and Brazil. It gathered insights from IT-decision makers from various industries, including financial services, information and communications technology, manufacturing, and retail.
Last month, an industry report by Thomson Reuters showed that GenAI adoption in professional services has nearly doubled over the past year. It revealed that organisational usage of GenAI increased from 12% in 2024 to 22% in 2025.
Read more: GenAI spending to surge to $644bn globally in 2025
