In the Philippines, a large share of workers earns income that never arrives in a neat payroll system. Retail clerks, transport drivers, BPO employees with overtime, freelancers, and small business owners often keep earnings in envelopes, photos, handwritten notes, or scattered bank credits. Traditional lenders usually ask for payslips, tax forms, or stable employment records, so these borrowers can appear invisible even when they have real cash flow. The practical challenge is not whether income exists, but how to verify it quickly and reliably.
That is where a tool that reads messy financial records becomes useful. For lenders, it can reduce the time spent checking documents and help standardize decisions across large numbers of applicants. For consumers, it may mean access to working capital for inventory, equipment, or cash-flow gaps without needing a long formal employment history. The benefit is real: more people and small firms could be served by banks, fintechs, or microfinance institutions. But the risk is also clear. If an AI system misreads a crumpled slip, ignores seasonal income swings, or pushes too much credit toward fragile households, it can create debt problems rather than solve them.
The regulatory backdrop matters. The Bangko Sentral has long pushed digital banking and financial inclusion, while the Data Privacy Act requires that personal financial information be collected with consent and protected appropriately. AI underwriting is not a separate license in itself; lenders remain responsible for fair lending, risk management, and compliance. Watch whether such systems are used by regulated institutions, how they explain decisions to borrowers, and what safeguards exist when documents are incomplete or ambiguous.
The next test is scale and performance. Early traction is one thing; whether repayment rates stay healthy as conditions tighten or informal income becomes more volatile is the harder question. Expansion into other proof-of-income sources, such as bank statements, remittance records, or supplier invoices, would show whether the model is broad enough for the Philippine economy. If it works well, this kind of technology could make credit less dependent on formal employment and more responsive to how Filipinos actually earn.