How does LeasePlain assign risk scores?

Risk scores (high, medium, low) are assigned based on a combination of factors: whether the clause may violate applicable law, the magnitude of potential financial impact on the tenant, how far the clause deviates from standard residential terms, and whether the clause restricts rights the tenant might otherwise have. The AI model is trained to apply these criteria consistently.

Are risk scores guaranteed to be correct?

No. Risk scoring is AI-generated and reflects general patterns in residential tenancy law. It cannot account for recent legislative changes, local court interpretations, or the full context of your specific situation. Use risk scores as a starting point for your own review and professional consultation, not as a definitive legal assessment.

What if my lease has a clause the AI misreads?

AI can misread poorly formatted text, miss page breaks, or fail to understand context that a human reader would catch. If an analysis seems wrong for a particular clause, trust your own reading of the lease and verify with a professional if needed. You can always re-upload a cleaner copy of your document.

Does LeasePlain's analysis change if the same lease is uploaded twice?

Slightly. Large language models have some inherent variability. Two analyses of the same document may use different phrasing or emphasize different aspects, but substantive findings should be consistent. If you get significantly different results, contact us — it may indicate a document formatting issue.

How does LeasePlain handle jurisdiction differences?

The AI model is trained with awareness of provincial and state tenancy legislation. When analyzing a lease, it considers the jurisdiction specified in the document. If no jurisdiction is clear from the document, it may apply general principles. For jurisdiction-specific accuracy, ensure your document clearly identifies the province or state.

LeasePlain Methodology

How we evaluate lease clauses, assign risk scores, and structure our analysis — and where the limits of AI-powered review lie.

How the Analysis Works

When you upload a lease, LeasePlain processes it through several stages: OCR extraction (for scanned documents), clause identification, plain-English interpretation, risk assessment, and structured output generation.

The AI model powering this analysis is a large language model (LLM) trained on legal documents, tenancy legislation, and residential lease templates from Canadian provinces and US states. It identifies clause types, extracts key terms, interprets their meaning in context, and flags items that deviate from standard practice or that carry tenant risk.

Our analysis is structured around six categories designed to give you a complete, actionable picture of your lease — from the basic terms to the negotiable details.

Risk Scoring Framework

Individual clauses are assigned one of three risk levels. Here is what each means:

High Risk

Clauses that could expose tenants to significant financial liability, restrict fundamental tenant rights, or that may violate applicable tenancy legislation.

Examples:

Clauses waiving the landlord's maintenance obligations
Automatic rent increases significantly above guideline amounts
Personal liability clauses that exceed what law allows
No-subletting clauses where subletting rights are legally protected

Medium Risk

Clauses that are legally permissible but require careful attention, impose meaningful obligations, or are commonly negotiated.

Examples:

Late fees that are significant but within legal limits
Strict notice requirements for repairs
Pet restrictions and associated fees
Limitations on guest stays

Low Risk

Standard, commonly used clauses that impose ordinary obligations and are unlikely to cause unexpected issues for tenants who understand them.

Examples:

Standard notice-to-vacate requirements
Routine maintenance responsibility assignments
Typical security deposit terms within legal limits
Standard quiet hours provisions

Six Analysis Categories

1. Lease Summary

Extraction and plain-English presentation of core lease terms: parties, property, term dates, rent amount, due dates, and renewal provisions.

2. Financial Terms

All monetary obligations: security deposit, last month's rent, rent increases, late fees, NSF charges, parking fees, utilities, and any other costs the tenant bears.

3. Red Flags

Clauses that are unusual, potentially unenforceable, significantly one-sided, or that deviate materially from standard residential lease terms in the relevant jurisdiction.

4. Unclear Clauses

Ambiguous, vague, or contradictory language that could be interpreted multiple ways — a common source of tenant-landlord disputes.

5. Questions to Ask

Specific, actionable questions the tenant should raise with the landlord before signing, derived from the actual content of the lease being analyzed.

6. Negotiation Points

Clauses commonly negotiated by tenants, with context on what changes are realistic to request and what concessions landlords typically offer.

Interpretation Framework

For each identified clause, the AI applies a consistent interpretation framework:

1
Identify clause type
What category of provision is this? (financial, maintenance, termination, occupancy, etc.)
2
Extract key terms
What specific obligations, amounts, dates, or conditions does the clause create?
3
Translate to plain English
What does this clause actually require of the tenant, in everyday language?
4
Compare to standard practice
Is this clause typical for the relevant jurisdiction, or does it deviate from standard lease terms?
5
Assess tenant impact
What is the potential financial, legal, or practical impact on the tenant if this clause is enforced?
6
Assign risk level
Based on the above, what risk level does this clause represent for the tenant?

Known Limitations

We are committed to being transparent about what our methodology cannot do:

The AI cannot account for recent legislative changes after its training cutoff date.

OCR errors in scanned documents can cause misreading of key terms.

The AI may miss context that spans multiple sections of a long lease.

Risk scoring reflects general patterns and cannot replace jurisdiction-specific legal advice.

The AI does not know your personal circumstances, rental history, or specific housing situation.

Analysis quality depends significantly on the readability and formatting of the uploaded document.

These limitations are why LeasePlain is a starting point, not an endpoint. Use our analysis to get oriented and identify what to investigate further — then verify important findings with a qualified professional.

LeasePlain Methodology

How the Analysis Works

Risk Scoring Framework

Six Analysis Categories

1. Lease Summary

2. Financial Terms

3. Red Flags

4. Unclear Clauses

5. Questions to Ask

6. Negotiation Points

Interpretation Framework

Known Limitations

Methodology Questions

See the methodology in action