AI And Legal Practice – In the Future Law Firms
- AI expected to reduce routine legal work
- Malcolm estimated 20–30% of mundane work could move to AI and legal-tech systems.
- Potential impact on junior staff, article clerks, training opportunities, and billable-hour models.
- Law firms may become leaner, more specialised, and more niche-focused.
- AI law identified as emerging practice area
- Potential work involving hallucinated cases, incorrect AI outputs, hacked systems, and accountability for developers.
- Possible dedicated AI-law departments within firms.
- No formal implementation decision recorded.
- Human oversight retained as operating principle
- D supported AI for research, large information sets, diarising, filing, and other mundane tasks.
- AI should enable legal practice rather than replace professional judgment.
- Outputs require checking, patrolling, and policing.
- Daily AI learning encouraged
- Malcolm recommended setting aside 15 minutes per day to follow AI developments or build useful agents.
- Suggested sources included Tech4Law, its AI For Lawyers section, forums, and recent YouTube content.
- F requested practical guidance and was directed to start with Tech4Law’s existing material
Legal Technology And Estates
- Existing legal-tech tools remain relevant
- Malcolm cautioned against replacing established systems solely because AI is now the dominant buzzword.
- Legacy platforms are incorporating AI rather than becoming obsolete.
- Xpressdox cited as an established document-assembly system with AI functionality.
- Estate-management software options identified
- D has two estates underway and expects one to be significantly more difficult.
- D asked for AI or legal-tech assistance beyond LexisNexis.
- Malcolm recalled a Pretoria or Centurion provider focused on estates but could not identify its name (Libralex link below notes.)
- LexPro presented as potential estate solution
- C described LexPro’s estate offering as strong based on a previous demonstration.
- Per-estate pricing was viewed positively; C recalled approximately R900 but flagged uncertainty.
- Local providers and deceased-estate discussion groups suggested as additional research sources.
- D to investigate available providers
- C recommended searching local providers and reviewing Facebook deceased-estate groups.
- D confirmed intention to conduct further research.
- No specific provider was selected.
Open Source And Infrastructure
- Open-source law-firm stack remains technically possible but costly to maintain
- Microsoft Office, Teams, SharePoint, Adobe, and similar tools were identified as major licensing dependencies.
- Alternative office and PDF applications could reduce licence costs.
- Compatibility, migration effort, and ongoing maintenance may offset savings.
- Microsoft dependency creates integration constraints
- K noted that a firm with approximately 120 users would require substantial investment to move fully into Microsoft’s legal and productivity stack.
- Existing independently hosted mail infrastructure would need integration.
- Some legal applications still depend on Microsoft products, pushing firms back toward Microsoft licensing.
- Digital sovereignty and uptime raised as open-source drivers (21:43)
- K identified concerns about dependence on one provider and the effect of Microsoft outages.
- Open-source options could provide greater control and potential cost benefits.
- Reliability and continuous firm uptime remain requirements.
- Chromebooks viewed as limited alternative
- Chromebook pricing was stated as approximately R4,000 to R12,000.
- Google Workspace could run in the Chrome OS environment.
- Microsoft-dependent legal applications and integrations remain a major limitation.
Billing And Client Transparency in the World of AI
- AI usage proposed as transparent client-enabled work
- Malcolm suggested informing clients during matter intake when AI will be used.
- Clients could approve AI use before work begins.
- Lawyers would continue billing for reviewing, analysing, and applying professional judgment to AI outputs.
- Productised legal services may expand
- Standard documents, such as sale agreements, could be offered at fixed prices through firm websites.
- Additional work could revert to hourly billing.
- Results-based or outcome-based billing was identified as another possible direction.
- AI consumption may become matter disbursement
- Token usage from AI agents could be recorded against individual matters.
- Firms may pass AI-related token costs to clients as disbursements.
- Hourly billing was expected to remain but potentially alongside product, outcome, and token-based charges.
Governance And Next Steps
- Firm-wide AI usage policy recommended
- Malcolm recommended a written policy for firms with staff.
- Policy should address improper or uncontrolled AI use and related litigation risk.
- Suggested starting point: identify the most AI-literate person in the firm to prepare a first draft, then tailor it internally.
- AI learning resources identified for follow-up
- F to begin with Tech4Law’s AI For Lawyers content. (Link below)
- YouTube searches and discussion forums suggested for current developments.
- Malcolm encouraged bringing relevant findings back for discussion.
Link spoken about and now shared:
Libralex, for “D” on estate systems, these guys have been around for years.
https://www.libralex.co.za/Home/#services
Tech4Law AI for Lawyers section:
https://www.tech4law.co.za/courses-on-offer/ai-for-lawyers/








