How to use AI responsibly - The Manila Times

TEXT START: ARTIFICIAL intelligence (AI) has begun influencing business decisions, strategic planning and daily operations.

The Dissection

The article converts AI’s structural threat into a governance-and-certification problem. It acknowledges job displacement, but quickly subordinates it to privacy, bias, deepfakes, accountability and stakeholder confidence. ISO/IEC 42001 is presented as the managerial mechanism that makes AI adoption appear controllable, auditable and internationally respectable.

The Philippine BPO sector’s projected $42 billion in export revenue and 1.97 million workers are framed as evidence of AI-enabled growth. Under the Discontinuity Thesis, they are also exposure data: a large concentration of routine cognitive labor is a concentrated automation target. The article treats that vulnerability as background noise while foregrounding certification.

The Core Fallacy

The central error is category substitution. The article treats the consequences of AI as failures of governance rather than consequences of successful automation.

ISO certification can reduce privacy breaches, improve audit trails and expose biased decisions. It cannot preserve the mass employment-to-wage-to-consumption circuit once AI performs cognitive work more cheaply and reliably than humans. It can make displacement orderly, documented and legally defensible. It cannot make displaced labor economically necessary again.

The article’s idea that “trust” determines whether AI progress is sustainable also reverses the causal hierarchy. Firms do not need universal trust to automate when competitive pressure rewards lower costs and higher throughput. Governance will be adopted largely because it protects market access, limits liability and reassures clients—not because it prevents productive participation from collapsing.

Hidden Assumptions

• Productivity gains will coexist with durable mass employment rather than removing the need for it.
• Better rules can constrain competitive incentives strongly enough to preserve human-only economic domains.
• Accountability assigned to managers or auditors remains meaningful when the underlying system’s objective is labor substitution.
• Stakeholder confidence will translate into broadly shared gains instead of legitimizing concentrated ownership of AI capital.
• Certification will function as substantive safety rather than as a procurement signal, reputational shield and compliance expense.
• Employees displaced by automation can be absorbed into new roles at sufficient scale and speed.
• “Responsible” AI is defined mainly by how safely organizations deploy it, not by who owns the productive system and who loses bargaining power.

None of these assumptions addresses the hardened sequence: cognitive automation dominance, inability to maintain human-only domains at scale, and collapse of economically necessary labor.

Social Function

Primary classification: transition management and ideological anesthetic, with elements of prestige signaling and partial truth.

The article is not worthless. Its warnings about privacy, bias, cybersecurity and unchecked automated decisions are operationally real. But its social function is to make an irreversible power transfer sound like a standards exercise. ISO 42001 supplies a respectable wrapper around the deeper event: firms are building systems that can reduce dependence on human workers, while governance language reassures clients, regulators and employees that the transition remains under human control.

This is managerial lullaby, not because every recommendation is false, but because the relevant truth is deliberately kept at the level of process. The article asks whether AI is governed responsibly. The harder question is whether a society can remain stable when the owners of AI capture productivity while the majority lose access to economically necessary work. That question is absent.

The Verdict

A competent compliance article describing the plumbing around a structural rupture. Responsible AI governance may reduce collateral damage and make automation easier to trust, but it does not alter the Discontinuity Thesis. It is hospice administration for the labor circuit: cleaner records, better safeguards and a more orderly decline.