41% of businesses find it difficult to integrate AI into existing systems

Research on the AI ​​development roadmap at enterprise scale was conducted by FPT and Forrester Consulting with 397 business and technology leaders in North America, Europe, Asia - Pacific and Japan, combined with in-depth interviews with senior leaders in many fields such as finance, manufacturing, automobiles, healthcare, energy and sports.

The results show that AI is creating productivity and operational efficiency benefits, but is still in the early stages of transformation. Among them, the biggest investment driver is increasing automation, with 48% choosing, followed by improving the ability to exploit data for decision making (45%).

Businesses that have implemented AI note a number of clear benefits, such as faster, consistent decision making and improved operational efficiency. However, AI is still mainly used to optimize existing processes instead of redesigning operating models. Only 34% said they are prioritizing the transition to an AI-first operating model, reflecting a significant gap between adoption and comprehensive AI-based transformation.

"It shows that AI has moved beyond technological experimentation to become a strategic business priority, but the transformation process is still in the early stages," the report said.

However, the research also shows that the biggest challenge lies in the enterprise platform instead of AI technology. Data, systems and governance models become the main bottlenecks when up to 41% of surveyed leaders said the biggest difficulty is integrating AI with existing systems and processes, 38% said that the ability to expand AI from pilot projects into a common platform for businesses is a big challenge.

Although more than half of businesses spend at least 5% of their information technology budget on AI, research shows that only 26% consider themselves to have reached a high level of operational maturity. On the contrary, expectations for AI agents are increasing rapidly. This tool currently covers an average of 17% of a business's core processes, expected to increase to 26% in the next 12 months and 39% in the next two years.

When deploying on a large scale, most said the criteria for choosing partners also changed from simply looking for technology suppliers to prioritizing partners with the ability to accompany throughout the AI ​​lifecycle, with platform cooperation models (62%), operational management services (58%) and co-creation (50%). Regarding investment priorities in the future, respondents said they will improve AI governance and compliance capacity (44%), integrate AI into existing systems (42%), and enhance security and resilience of AI systems (39%).

In addition, the research also shows that the next phase of AI is no longer a matter of testing individual applications, but instead aims to build an AI platform that can be reused and expanded across the company. According to Forrester Consulting, to realize this goal, businesses need a synchronous approach, combining technology infrastructure, data, administration, security and a unified operating model.

"As the AI ​​ecosystem becomes increasingly complex, businesses need partners with the ability to connect strategy, integration, management and operations to transform innovation into execution and replicable capabilities," commented Mr. Pham Minh Tuan, Deputy General Director of FPT Corporation and General Director of FPT Software. "Accompanying throughout the transformation journey helps businesses expand AI more quickly, safely, effectively and sustainably, thereby being ready for the next stage of development."

The 2026 Business Outlook Report published at the end of last month by UOB bank also shows that AI is becoming an important growth driver. Although most are still in the early stages of digitalization, about 80% of units have begun to apply AI to business operations, and 1/4 of businesses have implemented it at an advanced level. However, the process still faces challenges of investment costs and financial constraints (49%), lack of complete AI solutions and partner ecosystems (48%), or data and technology system readiness (47%).

Earlier this year, research published by AWS showed that by 2025, the rate of AI application in Vietnamese businesses will increase by 39% compared to 2024. 18%, equivalent to nearly 170,000 businesses, have deployed AI. Among businesses that have applied, 61% recorded an increase in revenue, with an average increase of 16%, while 58% expect AI to save about 20% in costs.

A number of units currently provide total AI solutions for businesses, from readiness assessment, roadmap building, platform development, system integration, administration to operations and scale expansion. For example, FPT develops CASAN to support businesses in assessing AI maturity levels and determining appropriate transformation roadmaps; or the FleziPT AI platform helps companies build and expand AI applications.

Bao Lam