Singapore businesses are charging ahead with artificial intelligence (AI) agents, but a stark divide is emerging between corporate ambition and employee acceptance, according to recent reports by Workday and Cisco.
Although 79% of organisations in Singapore are already deploying AI agents, workers are setting firm boundaries on how far they will let the technology into their professional lives. Most are comfortable working with AI as a collaborator, but only 8% would accept being managed by one, according to Workday’s report titled AI Agents Are Here – But Don’t Call Them Boss.
The disconnect highlights a critical challenge for businesses eager to harness AI's productivity gains, as employees want AI as a co-pilot instead of a commander. While nearly 90% of workers believe AI agents will help them accomplish more, half worry about a decline in critical thinking, and a third fear reduced quality in human interaction.
"We're entering a new era of work in Singapore where AI is an incredible partner to organisations today, complementing human judgement, leadership, and empathy. To drive productivity and trust, it is important that we rely on AI as a partner rather than a leader," says Jess O'Reilly, general manager for Asean at Workday.
The hesitation extends to autonomous operations, with just 24% comfortable with AI agents working in the background without human knowledge. Trust also varies sharply by task. Half trust AI for skills development, but that plummets to 15% for hiring decisions, 23% for finance, and 13% for legal matters.
As AI agents take on bigger roles, calls for stronger governance are growing louder. All Singapore respondents in Workday’s survey say regulation is needed, with 57% supporting ethical guidelines set by developers and 48% backing strict human oversight.
See also: Pony AI and WeRide get Chinese regulator’s nod to list in Hong Kong
Infrastructure struggles
The enthusiasm for AI agents is colliding with weak technical foundations. According to Cisco's third annual AI Readiness Index, 38% of organisations in Singapore say their networks cannot scale for AI's complexity or data volumes, and only 11% describe their networks as flexible or adaptable.
Yet, 30% of Singapore organisations expect AI agents to work alongside employees within a year. This creates what Cisco calls "AI Infrastructure Debt", or the accumulation of compromises, deferred upgrades and underfunded architecture that threatens to erode AI value over time.
See also: Samsung shares hit fresh record after AI chip boom lifts profit
"As organisations now move toward deploying AI agents, their success depends on their readiness, discipline, and action," says Tay Bee Kheng, president of Cisco Asean.
The pacesetter advantage
A small group of leading companies, dubbed “Pacesetters” by Cisco, are outperforming their peers. The study found they are five times more likely to move AI pilots into production and 70% more likely to deliver measurable value.
These leaders distinguish themselves through systematic preparation. They have clear AI roadmaps, strong change management plans, and are investing heavily in new data-centre capacity within the next 12 months. Their disciplined approach also extends to security, with Pacesetters highly aware of AI-specific threats and equipped to control and secure AI systems.
“This year’s Cisco AI Readiness Index makes one thing clear: readiness leads to value. Across the board, we are seeing that AI-ready organisations— the Pacesetters in our research — prove this,” says Tay.
