Tax agency's algorithm flags data inconsistencies, not fraud, for 2026 filing

The Spanish Tax Agency's (Agencia Tributaria) 2026 tax season, kicking off in April for the 2025 tax year, isn't about hunting down intentional tax evasion. Instead, a sophisticated algorithmic system is designed to identify discrepancies in your declared data – a shift in focus that could surprise many taxpayers.

The algorithm's focus: spotting data mismatches

Emilio Baena, a former inspector with over a decade at the Agencia Tributaria, revealed the core of this system on LinkedIn: “Hacienda doesn’t look for fraud. It looks for inconsistencies.” This isn't the old image of auditors meticulously reviewing every return. The system operates by cross-referencing your information with a vast network of data sources.

Think of it as an anomaly detector. The algorithm compares what you declare against information from third parties – employer withholdings, bank interest reports, historical spending patterns, and sector averages. A sudden jump in income, for instance, or a discrepancy between your declared income and bank activity can trigger a review.

The data integration is extensive. The Agencia Tributaria processed 1.2 billion records in 2025, highlighting the scale of this operation. This year, the focus is on identifying risk profiles based on these inconsistencies, not on assigning blame.

Baena’s insights shed light on common pitfalls. For example, failing to report income from a side hustle or not declaring rental income can easily trigger an alert. Even seemingly minor inconsistencies, like a significant deviation from the average income in your profession or region, can flag your return for further scrutiny. A bank transfer that doesn't align with your declared employment could also trigger a review.

The impact is particularly pronounced for vulnerable groups. Self-employed individuals with fluctuating income, those receiving the Ingreso Mínimo Vital, or those with unreported rental income face heightened scrutiny. A simple mismatch in bank data can lead to requests for clarification or even penalties if not addressed promptly.

The good news? A staggering 85% of returns are approved without changes. However, inconsistencies account for 70% of the limited checks. Understanding how the system operates – and reviewing your data carefully before filing – can significantly reduce your risk.

Baena emphasizes understanding how the Tax Agency thinks.