What you'll get with this service
Lexsys AI Document Assessment service starts with text extraction. The AI Legal Agent ingests the consumers finance agreement, whether it’s a PDF or Word file. It will also take in a scanned document and use OCR to extract the necessary text.
Pattern recognition
The AI scans for clauses historically flagged as abusive, e.g., excessive interest rates, hidden fees, unilateral modification rights.
Risk scoring
Lexsys AI assigns a risk level to each clause, e.g., “highly abusive,” “potentially abusive,” or “low risk,” based on regulatory benchmarks.
Contextual interpretation
Our AI assesses meaning, such as whether a clause effectively deprives the consumer of funds or essential rights.
Clause summaries
The Document Assessment summary generates plain language summaries of abusive or potentially unfair clauses contained in the uploaded agreement.
Frequently asked question
In practical terms, for a consumer claiming against abusive clauses in finance agreements, AI can reliably:
- Quickly identify high-risk clauses.
- Summarize potential legal issues.
- Support lawyers with evidence for filing claims.
A case is only valid if it is presented with the following:
- An Apud Acta, granting us Power Of Attorney, signed using the Lexsys provided mobile application, requiring your client to provide their unique government issued Digital Certificate and DNI number.
- An image of the front and rear of their DNI card. The Legal AI Agent will retrieve the DNI number from the card and match it to the DNI number on the provided Apud Acta.
- A copy of the Finance Agreement containing a claim which has been assessed by our Legal AI Agent Case Assessment system, and a receipt of at least one payment made against this finance agreement by the client. The DNI on the Finance Agreement is again retrieved and must match that on the provided DNI card and Apud Acta.
In practical terms, AI Legal Tech can detect between 10 and 15 distinct abusive clause types in a single standard consumer finance agreement.
AI legal tech relies on machine learning based on examples of contracts and annotated clauses. To learn a new claim type the system needs annotated examples, contracts with the new clause type clearly labeled and historical claims or successful court cases as training references.