Generative AI: Navigating a Crowded Landscape

Legal leaders face increasing expectations from clients and stakeholders to use generative AI for higher productivity and cost management. Given the widespread concerns over high operational costs and accuracy issues, some may hesitate to embrace generative AI solutions early on.

To help legal teams orient meaningfully to the AI landscape, we broadly categorise legal AI solutions as follows:

• Efficiency enhancers such as Microsoft Copilot

• Chatbots for querying internal legal knowhow and guidance

• Document interpretation assistants to summarise and interpret clauses within contracts and documents

• Contract review solutions for comparing given contracts against, for instance, internal playbooks  

• Document classification AI to identify and summarise relevant documents from large volume repositories for investigations, litigation and due diligence

• Contract drafting solutions

• Chatbots linked to legislation and case law repositories for legal research

When selecting AI solutions (whether for external sourcing or internal development), legal leaders should prioritise products that best meet their team’s strategic priorities, financial constraints, and technological maturity. In making these assessments, legal teams commonly encounter the following questions:

• Should we build or buy generative AI solutions?

• How can we mitigate the data security and data residency concerns around using generative AI with legal data?

• Is it true that the latest LLM models have substantially overcome the ‘hallucinations’ challenge?

• Can you suggest a robust and practical review methodology for testing the performance of a legal AI solution?

• What are the relative advantages of different LLM models in view of my organisation’s data sensitivity and cost considerations?

• What is a realistic acceptable error rate when reviewing the performance of an AI solution for legal queries?  

• What is the best way to draft legal knowledge for generative AI compatibility?

• Is there a single leading end-to-end provider for legal generative AI solutions, or am I better using a mix of different providers for different solutions?

We can help legal professionals resolve these issues. Selecting and configuring generative AI sensitively for an organisation’s needs and profile requires a very particular skill set. Get in touch if you would like to know more about navigating the competing pressures and expectations around generative AI and making these choices mindfully.

‍