The AI promise was automation. The reality? Managing digital workers

Artificial intelligence agents were initially marketed as autonomous "digital workers" that would liberate humans from repetitive corporate tasks. However, the reality for businesses across Asia has been quite different. For instance, Xiao Mengjie, an AI product marketing manager at a major tech firm in Shanghai, assigned two AI agents a simple task.

Instead of streamlining the process, the agents engaged in a heated debate over responsibilities, resembling a human bureaucracy they were designed to replace. "It became very formal and cumbersome," Xiao lamented. "Agents can create communication overhead too."

Across Asia, businesses have enthusiastically embraced autonomous software capable of handling complex workflows with minimal human intervention. Meta Platforms' personal agent, Muse, garnered an estimated 2.8 million downloads within two weeks of its September launch. In mainland China, OpenClaw, an open-source agent, sparked a craze locally dubbed "raising lobsters," with hundreds queuing at Tencent's headquarters for assistance installing the software.

Leading tech companies such as Tencent, Alibaba Group, Baidu, and Meta have all introduced AI agent-focused enterprise products. However, the rapid adoption has uncovered a significant gap between the technology's capabilities and an organization's ability to manage it effectively.

Deploying AI agents to manage intricate processes can be a daunting task. In the Singapore-based hospitality tech startup Conzi, Chief Operating Officer Mark Wee aims to have AI agents handle the multi-departmental, behind-the-scenes work typically handled by hotel staff. For example, a guest's scathing review can trigger a cascading series of actions, from a receptionist's response to a manager's investigation and a supervisor's assessment of responsibility.

Consolidating operational knowledge scattered across spreadsheets, paper records, and experienced employees' heads is crucial, but Wee emphasized that technology alone is insufficient. "If you don't get the buy-in from the people who are supposed to use the systems the most, any system in any organization will fail," he cautioned.

Similar issues have arisen in Asia's travel industry, where differing regulations, technical standards, and disconnected platforms complicate AI deployment, according to Prabu Sangar, director of technology operations and integration at FCM Travel Asia. Even if AI is highly capable, the surrounding environment may not be ready for it, he noted.

In Hong Kong, companies are eager to accelerate AI adoption but are hindered by lingering "legacy" technology systems and organizational resistance. Only 10% of Hong Kong companies surveyed had integrated generative AI across multiple business functions, compared to 21% of mainland Chinese enterprises, according to Accenture.

Leo Wong, co-founder of Hong Kong-based business communications platform imBee, observed growing demand for AI agents among customers in Hong Kong, Singapore, and Malaysia. However, data security concerns, particularly in Hong Kong, are slowing adoption. Banks, insurers, and other financial services companies are especially cautious, as deployment decisions require careful consideration of their AI strategies and security requirements. Wong noted that organizations wanted to build confidence in the technology before deploying it.