BSP Sets New AI Rules for Philippine Banks
On June 24, 2026, the Bangko Sentral ng Pilipinas (BSP) issued Memorandum No. M‑2026‑031, introducing the “Governance Principles for Artificial Intelligence in Financial Services.” Signed by Deputy Governor Lyn I. Javier, the guidance applies to all BSP‑supervised financial institutions (BSFIs) and their outsourced service providers.
The STARS Framework: Five Core Principles
The framework is built around five principles known as STARS: Sustainability, Transparency, Accountability, Responsibility (Social Fairness), and Security.
The Responsibility principle is particularly significant for customer service. It requires institutions to guard against biased training data, respect data privacy rights, and ensure AI systems do not exclude vulnerable groups from financial services. BSP warned that without proper safeguards, “AI systems may perpetuate biases, leading to unfair practices and the exclusion of individuals from access to financial products and services.”
While voluntary, the BSP made clear this represents its minimum supervisory expectations.
The Problem: Bias Happens in Real Time
The BSP’s warning is not abstract. In a contact center, bias can surface in subtle ways—a rushed agent providing incomplete information to a caller with a heavy accent, or inconsistent service quality depending on a customer’s background. The risk is not always intentional; it is often a blind spot in service design.
How Instadesk Agent Assistant Prevents Bias
Instadesk’s Agent Assistant is an AI‑powered real‑time productivity tool designed to support agents during live customer interactions. It addresses the Responsibility principle by ensuring every customer receives consistent, high‑quality service—regardless of who they are or how they speak.
Key capabilities include:
Automated Form‑Filling & Summarization – Extracts key info and generates concise summaries automatically, eliminating manual work and maximizing agent availability
Real‑time Agent Assistance – Provides real‑time transcription, sentiment analysis, relevant knowledge, and suggested actions directly in the agent workspace
Proactive Intelligent Recommendations – AI‑driven suggestions for knowledge base, modules, and next steps based on real‑time context, transforming a passive repository into an intelligent empowerment engine
Intelligent QA & Real‑time Alerting – Customizable rules monitor interactions in real time, instantly notifying supervisors of anomalies and enabling takeover to safeguard quality
From Reactive to Preventive
Traditional quality assurance catches problems after they happen—through call recording reviews and post‑interaction audits. Agent Assistant flips this model. It intervenes during the conversation, ensuring:
1. Every agent follows the same standardized service procedure, regardless of experience level
2. receive complete and accurate information, not abbreviated responses
3. Supervisors are ed to potential issues before they escalate
The Bottom Line
BSP’s STARS framework demands that financial institutions prevent bias, not just detect it. Instadesk Agent Assistant is built for exactly this purpose—helping banks serve every customer fairly, consistently, and in real time.
