AI in Legal and contract drafting

Note: This article is just one of 60+ sections from our full report titled: The 2024 Legal AI Retrospective - Key Lessons from the Past Year. Please download the full report to check any citations.

C. Legal and Contract Drafting

Recent advancements in AI-powered legal tools have significantly transformed the landscape of legal drafting, including legal agreements.

Legal templates have been used since documents existed, but they are often bloated as lawyers are incentivised to add-in new provisions, and rarely (if ever) take them out. We expect active development of AI document generation to take an increasing proportion of market share from template usage, via the use of playbooks, and data-driven contextual document generation provision-by-provision.

Shorter documents with more tailored coverage, can still maintain market standards. We have already seen this with the popularization of 'privacy policy generators' following the 2018 EU GDPR. Some are, naturally, much better than others and should be used with appropriate caution.

Some more 'classic' document automation solutions do exist to help law firms and businesses to streamline and automate the creation of legal documents. These are heavily reliant on primitive guided 'wizards' and underlying templates, which, while reducing the time spent on completely manually drafting documents from scratch, do not provide an optimal outcome.

Key Areas of Development

For the sake of clarity, we see the key areas of Legal AI development within legal and contract drafting within a quadrant.

On the horizontal axis you have high volume - low volume, and on the vertical axis you have low ~customisation, high customisation.

A CLM offers a brilliantly optimized solution to, say, manage a repeatable high-volume process with a low customisation document (such as rider agreements and NDAs for deliveroo riders). The recipient doesn't have bargaining power so negotiation has also effectively been reduced to zero. Customisation is near-zero. It is a process that can effectively be managed with a glorified mail-merge.

The introduction of AI into this process would be merely one of oversight, error-scanning.

As we move into the two quadrants with high customisation, regardless of volume, we see cutting-edge Legal AI beginning to bear fruit, and yielding a new class of Legal AI.