Case Law Search

Case Law Search

Find and analyze court decisions and arbitral awards on a specific legal issue. This skill structures the search, selects the right source by jurisdiction, retrieves verified decisions, and formats them for practitioner use. It never fabricates citations.

When to use this

• Researching how courts have interpreted a specific statutory provision or contractual clause
• Finding authority on whether a legal position is well-established or contested
• Locating precedent for a memo, brief, or legal opinion
• Checking whether a principle is "good law" (still followed, or overruled / distinguished)
• Building a table of authorities for a court submission
• Advising a client on litigation risk based on how courts have decided similar disputes

Inputs

Research pipeline

Step 1 — Identify jurisdiction and issue

Confirm jurisdiction via [[router-jurisdiction-detector]] if not explicit. Restate the legal issue with precision — the more specific the issue, the more targeted the search and the lower the hallucination risk.

Step 2 — Select authoritative source by jurisdiction

Step 3 — Search and filter

Search using:

1. Key terms from the legal issue (exact statutory phrases, well-known principle names)
2. Party type identifiers (commercial entity, employer, consumer) to narrow to analogous fact patterns
3. Date filter (default: most recent 10 years; for foundational principles, extend further)
4. Court level filter (apex / cassation > appellate > first instance)

Do not retrieve more than 10 cases unless the user has requested a comprehensive survey.

Step 4 — Verify each case

Anti-hallucination rule: never include a case that has not been verified as existing in an authoritative source. If the database returns a partial match or the case cannot be confirmed, exclude it and note the gap.

For each candidate case:

• Verify the caption (parties' names), court, year, and docket match the source record
• Confirm the holding stated matches the decision text
• Check citator status if available (still good law? overruled? distinguished?)

Output structure

For each case retrieved, produce:

**[Party A] v [Party B]**
Court: [full court name]  |  Year: [YYYY]  |  Docket: [number if available]
Citation: [official citation]

Facts (2–3 sentences): [what happened, who is suing whom, over what]

Holding (1 sentence): [what the court decided on the specific issue]

Key principle (1–2 sentences): [the legal rule extracted from the decision, stated abstractly so it can be applied to other facts]

Pin-cite: [paragraph / page number where the principle is stated, if available]

Citator status: [still good law / distinguished in X / overruled by Y / no subsequent treatment found]

Relevance to user's issue: [1 sentence explaining why this case matters to the specific query]

If comparing to the user's situation, add:

Distinguishing facts: [factual differences that may limit the case's application here]

Anti-hallucination rules

1. Never invent case citations. If no real case is found, say so explicitly: "No verified precedent found in [jurisdiction] on [issue] within the search scope."
2. Never paraphrase a holding in a way that extends beyond what the text of the decision actually says.
3. Never treat a lower-court decision as binding — clearly label persuasive vs binding authority.
4. Never treat a case decided before a statutory amendment as authority on the amended provision without checking whether the amendment changed the applicable rule.
5. If confidence in a citation is below "verified," do not include it. See [[router-confidence-scorer]] cite-or-bust rule.

MENA-specific notes

Arabic-language court systems

KSA and UAE onshore courts publish in Arabic. English summaries, where they exist, are unofficial and may omit nuance. For high-stakes matters, obtain the Arabic original and have it reviewed by a native-language-qualified lawyer.

Sharia-law courts (KSA)

Saudi courts are Sharia-based. Precedent functions differently: the principle of binding stare decisis does not apply in the same way as in common law; the court exercises independent ijtihad. However, Sharia legal opinions (fatwa) from the Council of Senior Scholars carry significant persuasive weight. The Commercial Courts (2017) operate under more codified rules.

DIFC and ADGM: common law islands

DIFC and ADGM are common-law jurisdictions within UAE. Their courts expressly cite and apply English cases as persuasive authority, making English High Court and UK Supreme Court decisions highly relevant. A case decided by the DIFC Court of Appeal is binding on the DIFC Court of First Instance.

The DIFC–Dubai onshore relationship

The DIFC Courts and the Dubai onshore courts operate under a judicial-transfer protocol. A DIFC Court judgment can be enforced onshore and vice versa. However, the applicable substantive law differs, so a DIFC case applying DIFC Contract Law is not directly binding on Dubai onshore courts applying UAE Civil Code.

New UAE federal legislation (post-2021)

UAE has undergone significant legislative reform since 2021 (new Companies Law, Labor Law FDL 33/2021, new Personal Status Law). Cases decided before these reforms may be on different statutory text — always check whether the statutory basis for the decision has changed.

Limits and escalation

• MENA public case-law databases are incomplete. Unreported decisions (which may represent the majority of first-instance decisions) are not accessible without direct court filing.
• In KSA, many commercial arbitration awards are not published. The Saudi Center for Commercial Arbitration (SCCA) does not maintain a public awards database.
• If the user needs exhaustive research for a high-stakes matter (client opinion, court filing), escalate to [[research-deep-research-orchestrator]] and pair with [[research-recent-amendments-tracker]] to check whether cited statutes have been amended.

Related skills

• [[research-statute-lookup]]
• [[research-recent-amendments-tracker]]
• [[research-precedent-finder]]
• [[research-deep-research-orchestrator]]
• [[router-jurisdiction-detector]]
• [[router-confidence-scorer]]