We’ve expanded spektr AI to interpret incoming PEP and sanction alerts and assess whether each is likely a true match to the entity or unrelated. The system consolidates structured sources (e.g., government watchlists, financial records, organizational databases) with unstructured sources (e.g., media reports, public disclosures, social media content, scanned documents) to build context such as position history, known associates, and other identifying details.

It then returns a structured assessment with clear rationale and links to supporting evidence so reviewers can verify the basis for each conclusion without manually cross-checking numerous records. PEP hits are categorized by severity (for example, higher-profile roles versus lower-risk positions), and you can configure risk thresholds to align with internal policies - controlling when alerts are surfaced, escalated, or de‑prioritized.