HÀ NỘI — In the face of a rapidly evolving global economy, the integration of artificial intelligence (AI) and digital technologies is becoming crucial for businesses looking to identify risks and adapt quickly to ensure sustainable growth.

However, while large enterprises are increasingly adopting these technologies, experts have said small and medium-sized enterprises (SMEs) are encountering challenges in leveraging AI for their operations.  

Recent research by IBM, Forbes and McKinsey highlights a steady increase in AI adoption within business governance, with 33 per cent of companies using AI in 2022, 55 per cent in 2023 and 72 per cent in 2024. Enterprises are incorporating AI into various sectors, including semiconductors, livestock and agriculture, to improve efficiency and gain a competitive edge.

For example, Cricket One, a large protein producer based in Bình Phước, has deployed AI since 2019 to monitor cricket living conditions. Co-founder Nguyễn Hồng Ngọc Bích told thesaigontimes.vn that although her company's AI platform still in the machine learning phase, but it provides valuable insights, allowing more informed and less emotional decisions in managing their cricket farm.

Similarly, Vinamilk, one of the country's largest dairy producers, has also embraced AI to enhance its operations. The company's Director of Agriculture Trịnh Phương Nam, said AI has been instrumental in analysing animal health data, predicting disease risks and improving treatment quality over the past five years.

However, experts said, small businesses face significant hurdles in adopting AI. 

Hoàng Trí, an owner of a small food retail business in HCM City attributed the failure to leverage AI to the high costs involved, like chatbots and management software, which can reach VNĐ10-20 million per month.

For small businesses with limited resources, the benefits of AI are often unclear, and the tools may go underused, Trí told thesaigontimes.vn.  

MISA, an IT solution provider, said that 97 per cent of SMEs face common obstacles when implementing AI. Key challenges include difficulties in managing fragmented data, the rising costs of scaling AI tools and the lack of investment in centralised AI systems. 

Many businesses also apply AI only in some departments, such as accounting or HR, resulting in operational inefficiencies, according to the company. 

AI: a key to business survival 

Recognising the transformative power of digital technologies, especially AI, experts said that AI has become essential for businesses to survive and thrive in an unpredictable economic landscape. 

Dr Lê Duy Tân from the AIoT Lab VN Centre advised businesses that lack the resources to build AI systems from scratch to have a gradual approach. AIoT Lab VN specialises in innovative solutions at the intersection of AI and IoT. 

He recommended starting with AI solutions provided by tech companies, learning how to use them effectively before considering in-house development or outsourcing to specialised units for key products tailored to their business needs.

Tân also pointed out that some small businesses in Japan have successfully outsourced AI solutions, while others with limited budgets can turn to subscription-based services from large technology firms. However, he said that relying on such comprehensive solutions might not align with each business's unique needs and data.

The General Director of Misa Company, Lê Hồng Quang, warned that a common pitfall for businesses is attempting to deploy AI on a large scale without a clear roadmap. This approach can lead to risks, especially if employees fear AI might replace their jobs or the business lacks a consistent data system. 

Quang stressed the importance of investing in digital platforms that ensure proper data collection, storage, and processing to avoid these issues. 

Once businesses had a solid data foundation, they could begin automating processes such as customer service with AI chatbots, analysing customer data for personalised offerings, and optimising supply chains, he told the online newspaper. — VNS