TwinLadder logoTwinLadder
TwinLadder
TwinLadder logoTwinLadder
Back to Archive
TwinLadder Intelligence
Issue #6

TwinLadder Weekly

April 2025

TwinLadder Weekly

Issue #6 | April 2025

AI in Small Claims: The £7.50 Legal Letter That Could Change Everything

UK startup Garfield.Law prices legal letters at £2-£7.50. Disruption or disaster? Here's what makes this different—and what doesn't.

Before we dive in: Next month's big story is already written. Garfield.Law will be authorized by the SRA as the UK's first AI-only law firm. But this issue focuses on the model itself—because the regulatory approval isn't the story. The pricing is.

The Access to Justice Problem

Here's the math that breaks small claims court in the UK:

  • Small claims limit: £10,000
  • Traditional solicitor fees: £1,500-£3,000+
  • Recoverable costs: Limited fixed amounts under CPR Part 45
  • Economic reality: Hiring a lawyer to chase a £2,000 debt makes no financial sense

The result? UK SMEs lose billions annually to unpaid invoices they can't economically pursue. The court system exists; the access doesn't.

This isn't new information. But someone finally built a solution that addresses the actual barrier: cost.

Enter Garfield.Law

Garfield.Law is an AI-powered litigation assistant founded by Philip Young (ex-Baker McKenzie partner) and Daniel Long (quantum physicist turned CTO). Their pricing model:

Stage Garfield Fee What It Does
Polite Chaser £2 Initial debt recovery letter
Letter Before Action £7.50 Formal demand letter (CPR requirement)
Claim Form £7.50 Court filing preparation
Default Judgment £7.50 Application for judgment when no response
Full Trial Support ~£50 total Complete claim through hearing

Compare that to £1,500+ for a traditional solicitor.

The key insight: 80% of claims end at the letter before action stage. Most debtors pay when they receive a formal legal letter. They just couldn't afford to send one before.

Why This Model Matters

1. Recoverable Costs

Almost all of Garfield's fees fall under the recoverable limits in CPR Part 45. If the claimant prevails, these costs come from the defendant—not the client's pocket. Traditional solicitor fees? Paid out-of-pocket regardless of outcome.

2. Scalable Supervision

The platform isn't autonomous. Each action requires client approval. Named solicitors remain accountable under SRA rules. But the AI handles drafting, procedure tracking, and document generation—the tasks that drive traditional legal costs.

3. Defined Scope

Garfield only handles small claims debt recovery up to £10,000. No general advice. No complex litigation. No scope creep. This narrow focus is a feature, not a limitation.

The Hallucination Safeguard

Here's the critical technical decision: Garfield's AI cannot propose case law.

The system is built on domain-specific expert knowledge modeled to follow Civil Procedure Rules. But it deliberately excludes case law citation—the highest-risk area for LLM hallucination.

Lord Justice Birss, deputy head of civil justice, called this approach "absolutely at the core of what we can do for access to justice."

They didn't solve the hallucination problem. They sidestepped it by building a tool that doesn't need to cite cases.

The Regulatory Picture

The SRA's eight-month approval process examined:

  • Quality-checking processes for AI work
  • Client confidentiality safeguards
  • Conflict of interest protections
  • Hallucination risk management

SRA Chief Executive Paul Philip stated: "The first regulatory approval of an AI-based law firm is a landmark moment for legal services in this country. With so many people and small businesses struggling to access legal services, we cannot afford to pull up the drawbridge on innovations that could have big public benefits."

The approval creates precedent. Other AI-native legal service providers will follow this path.

What This Means for Traditional Firms

For small claims work: The economics just changed. If you're a solicitor charging £1,500+ for small claims matters, your clients now have a £7.50 alternative. This isn't theoretical disruption—it's live competition.

For mid-market practice: The model demonstrates that AI can handle procedural legal work when scoped appropriately. The lesson isn't "AI will replace lawyers"—it's "AI can handle defined workflows at dramatically lower cost."

For larger firms: The narrowly scoped, high-volume, low-margin work that partners deprioritize? AI-native competitors will take it. The question is whether that's a threat or a welcome pruning.

Tool Review: Small Claims AI Platforms (UK)

Comparing the emerging players in AI-assisted small claims court access

Garfield.Law

What It Is: SRA-regulated AI litigation assistant for small claims debt recovery

Best For: SMEs chasing unpaid invoices under £10,000

Strengths:

  • Pay-as-you-go pricing (from £2)
  • SRA authorization provides regulatory confidence
  • Integrates with Xero, Sage, QuickBooks, FreeAgent
  • Most fees recoverable under fixed costs regime
  • No case law generation (eliminates hallucination risk)

Limitations:

  • UK only (England and Wales)
  • Debt recovery only—no general legal advice
  • Requires accounting software integration for optimal use
  • Client approval required at each step (by design)

Pricing: £2 polite chaser, £7.50 letter before action, £7.50 claim form Rating: 4.5/5 for its intended use case

CaseCraft AI

What It Is: Platform automating small claims from drafting to trial

Best For: Claimants and defendants managing claims without legal representation

Strengths:

  • Built for non-lawyers with plain-English guidance
  • Handles both claimant and defendant cases
  • Clear step-by-step process
  • No subscriptions—pay per use

Limitations:

  • Not SRA-regulated (legal information, not legal advice)
  • Newer entrant with less track record
  • Success fee model (10% of successful claims)

Pricing: £15 setup + 10% of successful claim value Rating: 3.5/5 - promising but less regulatory clarity than Garfield

Traditional Solicitors

What They Are: You know what a solicitor is

Best For: Complex disputes, claims near £10,000 limit, situations requiring advocacy

Strengths:

  • Full professional liability coverage
  • Human judgment on strategy
  • Handle escalation beyond small claims
  • Regulatory certainty

Limitations:

  • Cost-prohibitive for most small claims
  • Slow turnaround for routine matters
  • Non-recoverable fees mean economic loss even when winning

Pricing: £1,500-£3,000+ per matter Rating: Depends - excellent for complexity, poor value for routine debt recovery

What's Working: Access to Justice Success Stories

Success Story #1: The £3,400 Recovery

Business type: Freelance web designer Problem: Client ghosted after project completion

Before Garfield: "I asked three solicitors. Cheapest quote was £1,800. For a £3,400 invoice, it made no sense. I'd written it off."

After Garfield: "£7.50 letter before action. Payment received within two weeks. I genuinely didn't believe it would work—but formal legal correspondence changes the dynamic."

Key insight: The barrier wasn't legal complexity. It was economic access.

Success Story #2: The Repeat Customer Model

Business type: Small plumbing company Problem: ~10% of invoices go unpaid annually (avg. £800 each)

Approach: Integrated Garfield with accounting software. Automated trigger: Invoice 60 days overdue → polite chaser. No response in 14 days → letter before action.

Result: "First year: recovered £12,000 we'd have written off. Garfield fees: under £100 total. We don't think about debt recovery anymore—the system handles it."

Key insight: At £2-£7.50 per letter, systematic debt recovery becomes economically viable for the first time.

Hard Cases: Where the Model Struggles

Hard Case #1: The Disputed Invoice

Scenario: Client argues invoice amount is incorrect due to alleged service failures.

Problem: Garfield handles debt recovery where debt is undisputed. Genuine disputes over service quality or contract interpretation fall outside its scope.

Limitation: "The system sent a letter before action for £2,500. Client responded with detailed complaints about our work. Garfield couldn't help me navigate the dispute—I needed actual legal advice."

Lesson: AI works for procedural tasks with clear outcomes. Disputed facts require human judgment.

Hard Case #2: The Cross-Border Complication

Scenario: UK business chasing invoice from EU client post-Brexit.

Problem: Garfield operates under English and Welsh civil procedure. Cross-border enforcement is different legal territory.

Limitation: "Got the letter before action, but client is in France. Turns out enforcing a UK small claims judgment in the EU is its own nightmare. The tool couldn't help with that part."

Lesson: Jurisdictional scope matters. AI tools trained on specific procedures don't translate across borders.

Hard Case #3: The Strategic Debtor

Scenario: Debtor knows the system—waits until claim is filed, then pays just before hearing.

Problem: Some sophisticated debtors use delay as cash flow strategy, knowing small claims court is slow.

User report: "They paid three days before the hearing, after holding my money for 8 months. I 'won' but my working capital was tied up for most of the year."

Lesson: Faster legal access doesn't solve commercial reality—some debtors will game any system.

Reliability Corner

The UK Regulatory Landscape

Unlike the US patchwork, UK AI legal services operate under unified SRA oversight. Key requirements:

Requirement How Garfield Addresses It
Quality control Named solicitors accountable for AI output
Confidentiality Data handling subject to SRA inspection
Conflicts Systematic conflict checking before engagement
Hallucination risk Case law generation disabled by design
Client communication Approval required before each action

UK vs. EU Regulatory Comparison

Aspect UK (SRA Model) EU AI Act
Approach Sector-specific, principles-based Risk-based, cross-sectoral
AI regulator No new regulator (existing bodies) AI Office under European Commission
High-risk classification Case-by-case assessment Defined categories
Legal sector guidance SRA compliance tips General high-risk system rules
Status Framework active Mandatory compliance Aug 2026

Key takeaway: The UK's sector-specific approach enabled faster Garfield authorization. The EU AI Act may create different requirements for similar services operating in Europe.

This Month's Reminder

Court statistics show 66% of small debt claims result in default judgment. Most defendants don't respond. The AI isn't making complex legal decisions—it's automating a process where the outcome is often predetermined.

Workflow of the Month: AI Service Evaluation Checklist

Use this when evaluating any AI-assisted legal service for your practice or recommending one to clients.

AI LEGAL SERVICE EVALUATION
━━━━━━━━━━━━━━━━━━━━━━━━━━━

SERVICE IDENTIFICATION
Service name: _____________________________
Provider: _________________________________
Jurisdiction: _____________________________
Evaluation date: __________________________

REGULATORY STATUS
□ Regulated entity? (SRA, FCA, etc.)
  Regulator: ______________________________
  Registration number: ____________________
□ Named qualified professionals accountable?
  Name(s): ________________________________
□ Professional indemnity insurance in place?

SCOPE ASSESSMENT
□ What specific legal task does it handle?
  _________________________________________
□ What tasks are EXCLUDED from scope?
  _________________________________________
□ Is the scope appropriate for AI (procedural, defined outcomes)?
  YES / NO / UNCERTAIN

HALLUCINATION RISK ANALYSIS
□ Does the system generate case law citations?
  YES (higher risk) / NO (lower risk)
□ Does it produce legal analysis or opinions?
  YES (higher risk) / NO (lower risk)
□ Is human review required before output?
  YES / NO
□ What verification does the provider perform?
  _________________________________________

ECONOMIC ANALYSIS
Total cost for typical matter: $_____________
Traditional alternative cost: $_____________
Cost differential: $_____________

□ Are fees recoverable if successful?
  YES / NO / PARTIAL
□ Hidden costs? (setup, per-action, success fees)
  _________________________________________

CLIENT APPROVAL MODEL
□ Client approves each action? YES / NO
□ Automated actions without approval? YES / NO
  If yes, which? __________________________
□ Can client override AI recommendations?
  YES / NO

DATA HANDLING
□ Where is client data stored?
  _________________________________________
□ Data retention policy?
  _________________________________________
□ Third-party data sharing?
  YES / NO
  If yes, with whom? ______________________

RECOMMENDATION
□ Suitable for use case: YES / NO / CONDITIONAL
□ Conditions/limitations:
  _________________________________________
  _________________________________________

EVALUATOR: _____________ DATE: _____________

Time investment: 15-30 minutes per service Why it matters: AI legal tools are proliferating. Due diligence protects both you and your clients.

Quick Hits

UK Market News:

Regulatory Updates:

Coming Next Issue:

  • SRA Approves First AI-Only Law Firm: The Full Story (May 2025 follow-up)

Ask the Community

Garfield.Law raises questions about AI-assisted legal access we're researching:

  1. For UK practitioners: Have you encountered clients using AI small claims tools? What's been their experience?
  2. For multi-jurisdictional firms: How are you advising clients on AI legal tool selection across UK/EU?
  3. Access to justice question: Does £7.50 legal access help clients—or does it flood courts with marginal claims?
  4. Would you use an AI service evaluation template for client recommendations?

Reply to share. Anonymized contributions welcome.

TwinLadder Weekly | Issue #6 | April 2025

Helping lawyers build AI capability through honest education.

Sources

Previous Issue
Issue #5
TwinLadder Nedēļas Apskats
Next Issue
Issue #6
TwinLadder Nedēļas Apskats