Contextual Risk Analysis, or Smart Lexicon, uses LLM-powered detection to identify thematic compliance risks in communications, moving beyond traditional keyword-based rules. This approach reduces false positives and provides more meaningful alerts for compliance users, especially in environments with large volumes of unreviewed alerts.
Contextual Risk Analysis works by using predefined prompts to identify and create an alert for the following risk category signals:
Promissory Statements: Language that guarantees a specific outcome, promises a future return, or uses definitive "will/shall" language regarding value or performance.
Securities Instruments: References to stocks, bonds, mutual funds, ETFs, options, futures, derivatives, brokerage accounts, crypto-assets, IRAs, 401ks, or tokens. Does not flag insurance, annuities, bank products, loans, or mortgages.
Politically Charged Statements: Political or policy-related language that could be interpreted as partisan, advisory, or influential.
Threats: Threats of violence or harm to the company, employees, or leadership. Does not flag criticism, insults, or legal threats.
This feature is off by default. To enable it, contact Support at support@hearsaycorp.com.
Once enabled, you can configure which category signals are enabled by hierarchy. To configure categories, click Supervision, then click the Compliance Agent tab.