Rethinking AI’s Role in Legal Practice.

Time to Rethink AI’s Role in Your Legal Practice

In July 2025, a federal judge fined the attorneys representing MyPillow CEO Mike Lindell $3,000 each, not for misconduct in court, but for relying on AI to draft legal filings riddled with errors. Judge Nina Y. Wang of the U.S. District Court in Denver found 30 separate mistakes in the documents, including misquoted case law and citations to non-existent decisions.

What prompted the scrutiny? A single, pointed question from the bench:
“Was this motion generated by generative artificial intelligence?”
Attorney Christopher Kachouroff answered candidly:
“Not initially… I drafted a motion, and then we ran it through AI.”

That admission and the resulting penalties sent a clear message to the legal community: AI may be a powerful tool, but attorneys remain accountable for the quality and accuracy of their work.

This guide by Davis Page offers legal professionals a comprehensive overview of AI adoption, protecting both clients and careers in their ever-evolving regulatory environment.

Understanding the AI Adoption Surge and Its Hidden Risks

Data from the 2025 Generative AI in Professional Services Report reveals that AI usage among legal professionals has nearly doubled from 14% to 26% in just twelve months, with large firms reporting even higher adoption rates.

This growth masks a troubling reality: most attorneys are using consumer-grade AI platforms without understanding the data security implications.

Today, most attorneys are familiar with general AI models, such as ChatGPT and Anthropic’s Claude, which they assume function like a legal research database or expert consultant. The reality is very different.

How AI Processes Your Legal Queries:

  • Breaks down your input into components called “tokens”
  • Matches patterns from training data that can be months or years old
  • Predicts responses using statistical probability rather than factual verification
  • Builds answers word by word without consulting legal databases
  • Uses web-scraped data rather than authoritative legal sources
  • Generates plausible-sounding text regardless of factual basis

Why This Creates Legal Compliance Problems:

  • No fact-checking occurs before responses are generated
  • Training data becomes permanent and influences all future outputs
  • Systems learn citation formats without understanding the legal meaning
  • Responses sound authoritative despite being statistical predictions
  • Legal accuracy is not a guarantee

These models were designed for the general public. They are not trained in any particular area of expertise, making them highly susceptible to hallucinations. Instead, legal professionals must look to specialised tools that were designed to:

  • Cite case law: reasoning models that are specifically trained on legal documents will have a limited but more accurate database to pull from
  • Ensure proper formatting: Those models, such as ProPlaintiff.ai or CoCounsel, have been trained on thousands of documents that ensure proper formatting
  • Assist in case outcome analysis: Those models are particularly trained on case outcomes and will provide more accurate estimates than general models. Understanding these technical realities is crucial for attorneys who wish to utilise AI tools without compromising client confidentiality or professional standards.

Legal-Specific AI Platforms

Recognising that general AI models pose a potential legal compliance risk raises the question of how legal professionals can utilise this technology ethically and safely. This is where legal-specific AI platforms step in.

Enterprise Security Standards

Legal-specific AI platforms address the fundamental security gaps that make consumer tools unsuitable for attorney use. These platforms typically provide HIPAA compliance for practices handling health-related information, SOC 2 certification to ensure proper security and confidentiality controls, and data residency controls that allow attorneys to know where information is stored and who can access it.

Business Associate Agreements establish proper legal frameworks for data handling, creating contractual protections that consumer platforms don’t offer.

Legal-Focused Functionality

All the AI capabilities discussed earlier—strategic case development, document analysis, and outcome assessment—are available through compliant platforms designed specifically for legal professionals, such as CoCounsel and ProPlaintiff

Implementing AI: A Measured Approach

Integrating artificial intelligence into legal practice presents both significant opportunities and responsibilities. In this guide, we examined the dual nature of AI in law: its potential to enhance the efficiency and accuracy of legal work, as well as the considerable risks it introduces concerning client confidentiality, professional liability, and ethical duties.

Here’s some advice for implementing it in your practice:

Start with Low-Risk Applications:

  • Document review for discovery proceedings
  • Contract analysis for due diligence projects
  • Research assistance for case preparation
  • These applications maintain direct oversight while demonstrating AI’s value

Establish Clear Protocols Before Implementation:

  • Define which types of matters are appropriate for AI assistance
  • Specify required levels of human review and verification
  • Create documentation standards for AI-generated work
  • Remember: AI enhances capabilities but never replaces attorney responsibility

Launch with a Pilot Program:

  • Begin with a small team and a limited scope of work
  • Identify potential issues and refine processes
  • Build institutional knowledge before broader implementation
  • Document successes and failures to guide future scaling

The use of AI in the legal space offers clear benefits, including speed, pattern recognition, and data analysis, which are particularly beneficial for lawyers and their clients. However, to truly maximise its benefits, we require robust safeguards, ongoing education, and a steadfast commitment to ethics. The future belongs to those who embrace change, striking a balance between optimism and a healthy dose of scepticism, rather than simply digging in their heels or blindly accepting it.

By 
Davis Page
ProPlaintiff.ai

Back to blog