AI Agents Must Be as Accountable as Human Workers: BT Opinion
Source: The Business Times
The era of treating AI agents as experimental pilots is over, argues a Business Times opinion piece. Singapore’s move to create an AI agent registry signals that autonomous AI tools have shifted from curiosities to operational realities that need to be tracked, owned and governed like human employees.
AI agents are becoming colleagues, not just tools. That’s the central argument of a Business Times commentary published Monday, which argues that the defining challenge of the agentic AI era is the shift from adoption to accountability. As Singapore builds its AI agent registry for 150,000 public officers, the piece contends that a government acknowledging the need to track and govern AI agents marks a fundamental change in how we think about artificial intelligence in the workplace.
The piece traces Singapore’s agentic AI journey through four stages: the Agentic AI Primer (education), the Agents Sandbox (controlled piloting), the governance framework (establishing the rules), and the registry (formal tracking and ownership). Each stage reflects a growing recognition that AI agents are qualitatively different from the chatbots that came before — they don’t just answer questions, they act. An AI agent with access to corporate systems is less like a calculator and more like a new hire with a laptop and system access but no onboarding.
The comparison is deliberate. If a human employee can book a meeting, delete files and send emails, an AI agent with those same permissions should face equivalent governance. The piece argues that Singapore’s approach — building the registry before scaling agent deployment — gets the order right. GovTech’s Agentic AI Primer, published in April 2025, already framed autonomous AI tools as requiring the same formal governance structures applied to human teams.
The article also points out that the risks surrounding AI agents are no longer theoretical. Errors now have direct financial and security consequences. For AI to function as an operational reality, the same tracking, ownership and responsibility frameworks that govern human workers must apply to their AI counterparts.
Why it matters for Singapore: This piece captures why Singapore’s governance-first approach to agentic AI matters beyond government IT. As more Singapore companies deploy AI agents in customer-facing roles — in banking, healthcare, logistics — the accountability question becomes commercial, not just bureaucratic. Companies that treat AI agents like tools rather than quasi-employees are the ones that will face the regulatory and reputational blowback.