Hold music is dying. So is the entire commercial logic that made it necessary.

For decades, customer service operated under a major constraint: companies couldn’t afford to hire enough humans to answer every call instantly. The solution? Make people wait. Queue them. Play that music. Hope they don’t hang up (or maybe hope they do).

AI agents are killing that trade-off. New data from Salesforce‘s State of Service report, based on a survey of 6,500 service professionals globally, including 100 in Singapore, shows AI is already handling 30% of customer service cases in the city-state. By 2027, that figure jumps to 41%.

“For decades, customer service has been limited by a commercial constraint: businesses couldn’t afford to hire enough staff to answer every call instantly, so they relied on workarounds like hold music to manage the volume of inquiries. AI agents eliminate this trade-off, solving for both scale and quality. Instead of rationing exceptional service, companies can now use AI agents to deliver the immediate, tailored attention of a personal concierge to the mass market,” says Gavin Barfield, Salesforce’s vice president and chief technology officer for solutions in ASEAN. “This allows human teams to stop managing queues and start managing the complex, high-value relationships that truly drive growth.”

For data leaders, this isn’t just a customer service story. It’s a fundamental shift in how organizations allocate cognitive labor. And it’s happening faster than expected.

The rise of the agentic enterprise

AI has rocketed up the priority list for Singapore’s service leaders, jumping from #9 to #3 in just one year. The reason? Results.

Service teams project agentic AI will boost upsell revenue by 15%. They’re betting on AI to cut costs, improve customer satisfaction, and free up human workers for higher-value tasks. This is what Salesforce calls the “agentic enterprise,” where AI agents work as collaborative partners, reasoning and acting independently while humans focus on complex problems.

The data shows this isn’t speculative. Service reps using AI in Singapore spend 20% less time on routine cases, freeing up roughly four hours per week. Those using advanced agentic AI dedicate a full quarter of their week to the thorniest, most complex issues.

That redistribution of time matters. Compared to non-users, AI-enabled service reps are significantly more likely to mentor colleagues, lead cross-functional projects, and work with high-value customers. They’re moving up the value chain.

Workers are optimistic (for now)

The more interesting part of the report is that 84% of Singapore’s service reps with AI say it’s creating growth opportunities. Three-quarters have developed new skills. 78% say their role has become more specialized.

These workers aren’t worried about being replaced. Rather, it’s the opposite: They’re excited about being upgraded.

“Most importantly, APAC service reps with AI feel good about where they’re headed, with agentic AI users being the most optimistic about their career prospects,” the report notes. Whether that’s selection bias among early adopters or genuine transformation, the sentiment reveals something crucial: AI adoption in customer service isn’t triggering panic. It’s creating perceived opportunity.

For data leaders rolling out AI initiatives, that’s valuable intel. Change management works better when workers see upside.

Security remains the speed bump

Still, implementation isn’t frictionless. Security remains the top concern, with 49% of Singapore’s service leaders saying security worries have delayed or limited their AI initiatives.

86% alsp say the obstacles they’ve faced were expected. Security leaders are increasingly seeing AI as part of the solution, citing improvements in threat detection, anomaly monitoring, and breach prevention.

The lesson for data leaders? Security concerns are real, but they’re manageable. And they’re not stopping the wave.

Why this matters now

Customer service is often viewed as a cost center rather than a strategic priority. But it’s one of the few domains where AI agents can demonstrably deliver value today and not in some distant future.

Singapore’s customer service teams are running a live experiment in AI-human collaboration at scale. The data shows it’s working. Costs are dropping. Revenue is growing. Workers are adapting.

For data leaders watching from other sectors, Singapore’s 41% figure is a benchmark. It’s proof that AI agents can handle significant cognitive workloads in production environments. And it’s a signal that the agentic enterprise is already here.

The hold music is fading out. The question is what you’re building to replace it.

Image credit: iStockphoto/ipopba