Mastering ligi.ai: A Lawyer's Guide to AI-Powered Legal Research

Legal research has evolved. Malta's first AI-powered legal research platform, ligi.ai, combines the comprehensiveness of traditional legal databases with the natural language capabilities of modern AI—but it works differently from both.

This guide will help you get the most accurate, relevant results from ligi.ai by understanding how to structure your searches effectively.

The Fundamental Principle: Specific Questions, Not Keywords

Traditional legal databases trained us to think in keywords: "contract breach damages." ligi.ai works differently.

Instead of keywords, write questions or describe scenarios

❌ Bad: "company charge failure registration consequences"

✅ Good: "If a company failed to register a charge on time, what are the consequences of this omission under Chapter 386?"

The difference? ligi.ai's AI understands context, legal concepts, and the relationship between facts and law. Give it enough context to understand what you're actually looking for.

Three Research Approaches for Maximum Accuracy

1. Fact-Matching: Tell the Story

When you're researching how courts have handled similar fact patterns, describe the scenario as a short legal problem.

Use scenario triggers

- "Suppose that..."

- "If a person..."

- "When a company..."

- "What are the consequences of..."

- "Is [party] liable for..."

- "What obligations does..."

Example:

"Suppose that a director signed a contract on behalf of a company without proper authority. According to case law, what are the consequences under Chapter 386?"

This approach works best when you're looking for precedents involving similar factual circumstances.

2. Statutory Specificity: Pinpoint the Law

When researching the interpretation or application of specific legislation, directly reference the Chapter and Article.

Use legislation triggers:

- Chapter, Kapitolu, Chp, Kap (to identify the law)

- Article, Artikolu, Art., sub-article (to pinpoint the provision)

Example:

"Find all cases discussing the application of Article 1127 of Chapter 16 regarding non-performance of contractual obligations."

Pro tip: Combining Chapter and Article keywords triggers ligi.ai's most stringent relevance check, delivering the most accurate results.

3. Conceptual Search: Legal Principles and Maxims

When researching the application of legal principles, use either the English translation or the original Latin maxim.

Example:

"Find judgments applying the principle of Nemo Judex in Causa Sua."

This works because ligi.ai understands legal concepts across both languages and can identify when courts have applied specific doctrines, even when they're expressed differently across cases.

Critical Guidelines: What ligi.ai Does (and Doesn't Do)

DO ask for:

✅ Specific cases addressing particular facts

✅ Judicial interpretation of statutes

✅ Application of legal principles in case law

✅ Lists of relevant precedents

DON'T ask for:

❌ General legal advice

❌ Definitions of legal principles

❌ Information outside Malta case law

Why this matters: ligi.ai is designed to minimize AI hallucinations by searching only verified case law. It won't provide general legal knowledge or invented answers—it searches actual judgments and returns verifiable results.

Language Considerations

Always prompt in English and expect responses in English.

ligi.ai handles multilingual case texts automatically. When you search in English, the system:

- Searches cases in both English and Maltese

- Provides analysis in English

- Shows case summaries in the predominant language of each judgment

You don't need to search twice or worry about missing Maltese judgments. The AI handles cross-lingual search automatically.

Frame for Results, Not Answers

Structure your prompts to elicit a list of relevant cases rather than a single answer.

❌ "What is the statute of limitations for breach of contract?"

✅ "Find cases discussing the statute of limitations for breach of contract under Maltese law."

The first seeks a general answer. The second will return actual case law you can cite and verify.

Using Filters Effectively

When you need to narrow results by court, date, or adjudicator, use the platform's filters instead of including these criteria in your prompt.

❌ Don't write: "Find recent Court of Appeal cases from the last 5 years about..."

✅ Instead: Write your substantive query, then apply the date and court filters.

Why? The AI model processes your prompt for legal concepts and facts. Including administrative criteria like "recent" or "Court of Appeal" in the prompt dilutes the substantive analysis. Use filters for administrative narrowing.

The filters allow you to specify:

- Court level (Constitutional Court, Court of Appeal, etc.)

- Date ranges

- Specific adjudicators

Putting It All Together: Example Research Scenarios

Scenario 1: Researching Contractual Liability

❌ Weak prompt: "breach contract damages"

✅ Strong prompt: "If a contractor abandoned a construction project before completion, what damages have courts awarded under Article 1133 of Chapter 16?"

Then apply filters: Last 10 years, Civil Court

Scenario 2: Statutory Interpretation

❌ Weak prompt: "company law directors duties"

✅ Strong prompt: "Find cases interpreting the fiduciary duties of directors under Article 136 of Chapter 386."

Then apply filters: Court of Appeal, Specific judge if researching their jurisprudence

Scenario 3: Precedent for Specific Facts

❌ Weak prompt: "employment termination unfair dismissal"

✅ Strong prompt: "Suppose that an employee was dismissed without notice after 15 years of service. What compensation have courts awarded in similar circumstances?"

Then apply filters: Last 5 years, Industrial Tribunal

Quick Reference: Prompt Formula

For fact-based research:

[Scenario trigger] + [Factual description] + [Legal consequence question] + [Chapter/Article reference if applicable]

For statutory research:

"Find cases" + [discussing/interpreting/applying] + [Article X of Chapter Y] + [regarding specific issue]

For conceptual research:

"Find judgments" + [applying/discussing] + [legal principle or maxim]

Common Mistakes to Avoid

1. Keyword thinking

Malta's legal research culture trained us in keywords. Resist this habit. Write complete questions.

2. Including dates/courts in prompts

Use filters instead. Keep your prompt focused on legal substance.

3. Asking for definitions

ligi.ai searches case law, not legal dictionaries. Ask where courts have applied concepts, not what they mean in theory.

4. Overly broad questions

"Tell me about contract law" is too broad. "Find cases where courts have awarded specific performance under Article 1314 of Chapter 16" is appropriately focused.

5. Mixing languages

Prompt in English, even when researching Maltese-language cases. The system handles translation automatically.

Verification: Your Professional Responsibility

Every case returned by ligi.ai links directly to the source judgment on ecourts.gov.mt. Always click through to verify:

- The citation is accurate

- The context matches your understanding

- The precedent is appropriately applied

- The case hasn't been overturned or distinguished

ligi.ai eliminates hallucinations by searching only real cases, but professional verification remains your responsibility.

Start Researching Smarter

Understanding how ligi.ai processes queries helps you extract maximum value from AI-powered research. The platform combines comprehensive coverage of Malta case law with sophisticated natural language understanding—but the quality of your results depends on the quality of your prompts.

Remember the core principles:

- Write questions, not keywords

- Use scenario, legislation, or conceptual triggers

- Combine Chapter and Article references for precision

- Use filters for administrative narrowing

- Always verify results against source judgments

Master these techniques, and legal research that used to take hours now takes minutes.

Ready to transform your research process? Log in to app.ligi.ai and put these techniques into practice.

---

Need help? Contact [email protected] or schedule a training session to learn advanced search techniques specific to your practice areas.