AI is a tool. Stop treating it like a strategy

There are AI-powered Swiss army knives, with a blade, a saw, scissors, a USB drive, and a laser pointer. These all-in-one devices can handle various tasks, but when you need a specific function, like tightening a screw, you reach for the appropriate tool. The same principle applies to AI in the startup and corporate world. Instead of asking "How do we use AI for this?" the more pertinent question should be "What outcome do we need, and what is the right way to achieve it?"

This is where most AI implementation mistakes occur. Design thinking requires starting with the desired outcome, existing constraints, and the target user. AI is undeniably a powerful tool, but it remains a tool, not a solution. Applying AI blindly creates impressive but ultimately ineffective implementations. In corporate travel, AI excels at data-rich, pattern-driven tasks like pre-trip supplier research and rapidly generating rebooking options.

However, much of what matters in corporate travel, particularly in Southeast Asia, remains unstructured and not digitized. This includes relationships with hotel managers, knowledge of travelers' preferences from past trips, and understanding supplier policies – information not stored in databases. When AI lacks the necessary context, it produces outputs that may be confident but not accurate.

Travellers recognize this, as reflected in adoption data showing that while they are comfortable using AI for inspiration and research, they are reluctant to let AI handle bookings without human oversight. The real issue lies in accountability: when AI makes a mistake, who is responsible? In professions like medicine, law, and accounting, clear accountability exists because decisions are consequential and someone must be held accountable.

Corporate travel, however, sits in a similar category, with crisis management being a prime example. Decisions regarding employee safety and duty of care cannot be reduced to retrieval tasks. An AI system cannot fully comprehend the nuances of complex situations involving medical appointments, client relationships, and coordination with colleagues.

The industry acknowledges this, but the responsibility gap remains unaddressed. The fundamental problem is not technological but rather a matter of judgment. People instinctively recognize that booking a trip involves context that a system cannot capture. Thus, the trust and accountability challenges in corporate AI-driven travel programs stem from a gap in understanding and addressing the human elements inherent in the process.