Responsible AI.
Human Oversight.
No Autonomous Legal Decisions.
GAAS LAW operates on a clear governance model: AI handles 70% of routine operational work, while lawyers retain the critical 30%—judgment, strategy, client advice, and legal decision-making. All AI outputs require human review. No autonomous legal decisions. Full audit trails.
The 70/30 Rule
GAAS LAW is designed around a fundamental governance principle: AI should handle routine operational work that does not require professional judgment, while lawyers retain control over the work that matters most—legal strategy, client advice, negotiation, and courtroom advocacy.
AI Handles 70%
Routine Operational Work
- Data extraction from uploaded documents
- Timeline generation from fragmented evidence
- Audiogram pattern recognition and age correction
- CLB/LCB/Moore methodology calculation execution
- Consistency checking across historical matter data
- Missing evidence flagging and chase reminders
- Structured document generation (chronologies, summaries)
Lawyers Retain 30%
Critical Legal Judgment
- Legal strategy and defence approach
- Client advice and communication
- Professional judgment on causation and liability
- Settlement negotiation and authority decisions
- Courtroom advocacy and representation
- Review and approval of AI-generated outputs
- Override and adjustment of AI recommendations
Human Oversight Requirements
No Autonomous Legal Decisions
GAAS LAW does not make legal decisions autonomously. AI-generated outputs—medical summaries, causation calculations, liability assessments, quantum estimates—are marked as "pending approval" until a qualified fee earner reviews and confirms them. The platform provides structured analysis and operational support, but legal judgment remains exclusively human.
Governance Rule: No AI output is treated as final until explicitly approved by a qualified solicitor or fee earner. All outputs display approval status and approver identity.
Approval Checkpoints
Critical workflows require explicit human approval before outputs inform defence strategy or settlement decisions. Medical review summaries, causation methodology calculations, and liability assessments must be approved by supervisors before they are used in client advice or negotiation.
Medical Review Approval
Supervisor confirms medical summary before it informs causation analysis.
Methodology Approval
Partner approves causation methodology before it informs settlement strategy.
Settlement Authority
Settlement above delegated authority requires partner approval.
Insurer Reporting
External reports require supervisor review before release to insurers.
Explainability & Transparency
Every AI suggestion includes rationale—what data informed the output, what patterns were detected, what assumptions were made, and what confidence level applies. Fee earners can review the reasoning behind every AI-generated summary, calculation, or recommendation before approving or overriding it.
Transparency Requirement: All AI outputs display: data sources used, confidence score, assumptions made, and alternative interpretations flagged. No "black box" recommendations.
Legal Privilege Protection
All matter data, AI outputs, internal notes, and operational intelligence generated within GAAS LAW are treated as privileged client work product. The platform does not share, export, or use client matter data for model training without explicit consent. Role-based access controls prevent unauthorized disclosure.
Privileged Data Handling
All matter data marked as privileged. No cross-matter data sharing without consent.
No Model Training on Client Data
Client matter data is not used to train AI models without explicit permission.
Full Audit Trails
Every action within GAAS LAW—AI output generation, human review, approval, override, export—is logged with full audit trails. Who approved what, when, and why. What assumptions were overridden and by whom. What methodology was chosen and who confirmed it. This provides SRA-compliant, procurement-friendly audit documentation.
Action Logging
Every approval, override, and export logged with user ID, timestamp, and rationale.
External Audit Readiness
Audit trails exportable for external review, regulatory compliance, or insurer audit.
Override & Adjustment Controls
Fee earners can override any AI-generated output and document their rationale. If the platform suggests a CLB methodology but the solicitor believes LCB is more appropriate, they can override the suggestion and explain why. All overrides are logged and visible to supervisors.
Professional Judgment Primacy: Solicitors always have the final say. AI provides structured analysis, but professional judgment overrides machine recommendations when appropriate.
Model Governance & Supervision
GAAS LAW operates with clear model governance protocols. AI models are version-controlled, outputs are monitored for consistency, and performance is reviewed regularly. Partners and supervisors receive regular reports on AI output approval rates, override frequency, and confidence score distributions.
Supervision Dashboard
Partners see AI approval rates, override frequency, and outlier detection.
Version Control
AI model versions tracked. Changes to methodology or outputs fully documented.
Built for Responsible AI Adoption
General Counsel
Clear governance model, human oversight requirements, privilege protection, and audit trails provide confidence that AI adoption does not compromise professional responsibility.
Risk & Compliance Teams
No autonomous legal decisions, full audit trails, explainability, and approval checkpoints ensure regulatory compliance and risk mitigation.
Managing Partners
Operational efficiency without sacrificing professional judgment. The 70/30 Rule ensures fee earners focus on high-value work while AI handles routine operations.
Responsible AI. Human Judgment. Full Transparency.
GAAS LAW provides operational intelligence without compromising professional responsibility. AI handles routine work. Lawyers retain control over legal decisions. Full audit trails ensure accountability.