History and today's challenges

In the rapidly evolving landscape of legal technology, the integration of generative AI tools presents both unprecedented opportunities and important ethical challenges. Ryan Groff, a prominent member of Massachusetts bars and lecturer in New England law, explores these dimensions in his enlightening webinar.The ethical use of generator AI in legal practice.“

In the webinar, Ryan Groff discusses the ethical implications of using generative AI (genai) in legal practices, tracks and distinguishes the history of genai applications in law. Various AI tools available today. He provides an insightful overview of the historical application of genai in legal contexts, distinguishing between the various AI tools currently available.

Jump to below

A brief history of AI in law

---

Ethical AI Framework

---

Implement AI legally and ethically

---

How to Choose an Ethical Legal AI Solution

A brief history of AI in law

Groff then provided a brief history of AI in the legal sector, Adjusted to meet specific needs A legal expert. Groff also described the various types of AI tools available today, and outlined the general AI tools and tools specifically designed for legal tasks to explain the wide range of technologies available to legal experts.

It is worth noting that these technologies are being developed within a broader regulatory framework, including EU AI law, ABA ethical opinions on technology, and various state bar association guidelines addressing the use of AI in legal practices. Groff emphasizes that while AI can increase the efficiency of legal practices, it should not undermine the critical judgment of lawyers. He emphasizes the importance of maintaining strict supervision, protecting client confidentiality and ensuring technical capabilities.

Major Legal Technology Milestones:

• 2012: Predictive coding is Da Silva Moorev. Get judicial approval at PublicisGroupe
• 2016: Ross Intelligence launches IBM Watson-driven legal research platform
• 2018: Contract Analytics Platform begins using machine learning for due diligence
• 2021: GPT-based legal drafting assistants appear
• 2023: Domain-specific legal LLMS with reduced hallucination rates
• 2024: Integrating legal reasoning capabilities and reliable source verification
• 2025: Deep search and AI agents will lift the industry and increase the impact of what AI can do

With its long history with AI, Thomson Reuters has consistently pushed the boundaries of what is possible. Thomson Reuters has developed tools that reflect the evolution of Westlaw Edge's predictive analytics, practical legal analytics, and Cocounsel's professional legal capabilities built on domain-specific models.

Especially now, with the rise in agent AI (not to be confused with generator AI), there are ways that legal experts can continue to save time and increase productivity. Agent AI allows a single prompt to trigger multi-step tasks. Agent AI can create a wider range of workflows than Genai, helping you with tasks such as running legal investigations, comparing draft documents, and preparing for deposits on a single project.

Ethical AI Framework

Basic assumptions about legal ethics and AI

Groff set the stage by asserting these five principles.

1. AI as an assistant, not an alternative: AI should strengthen the competence of lawyers and act as legal assistants rather than replacing expert judgements.
2. Lawyer's liability: Attorneys are ultimately responsible for verifying all AI output before relying on professionally.
3. Avoid unfair practices: Unsupervised AI use may constitute illegal practices on behalf of the law.
4. Maintain your ability: Attorneys need to understand AI capabilities and limitations for proper deployment.
5. Confidentiality: If you use a third-party AI platform, you must protect the confidentiality of the client.

I did not adhere to these Ethical Principles It could result in disciplinary action, medical malpractice exposure, and violation of fiduciary duties to the client. No matter what tools you use, it is important for lawyers to maintain independent judgment and commit to legal ethical practices.

Understanding responsible developments

Emphasizing the ethical development of AI, Groff emphasized the importance of user control and accountability, and advocated a full understanding of AI's capabilities and limitations to ensure proper use.

Evaluation of the Ethical AI Framework:

• Clarity: Can AI clearly explain its inference process?
• Accuracy: Is the output effectively legally correct?
• Explanationality: Can lawyers understand and explain how AI has reached its conclusion?
• Auditability: If I try, can I review and check the process?

Potential and Limitations of AI in the Law

Groff pointed out the key domain-specific engineering needs of AI tools and advised them to suit legal purposes. He also has the potential and AI pitfalls in legal practices.

Specific restrictions and meanings:

• Hallucinations: AI can generate plausible but incorrect information, especially in legal contexts where accuracy is most important.
• Data biasModels trained in historical cases could perpetuate existing biases in the legal system
• Lack of transparency: The “black box” feature can make inference difficult to understand
• Limited Contextual Understanding: AI may miss subtle client needs or jurisdiction-specific requirements

Mitigation Strategy:

• Loop Human Review: Implement the required human verification protocol
• Red Team ToolsSystematically test AI tools to identify vulnerabilities before deployment
• Limited Use CasesLimit AI to low-risk tasks until reliability is established
• document: Maintain a record of when and how AI was used in client issues

Implement AI legally and ethically

In addition to understanding ethical principles, legal experts need practical guidance on AI operation in their practice. This is the framework for implementing law:

Step-by-step AI integration strategy:

1. Evaluation stage: Identify the specific use case for which AI will add values
2. Test phase: Pilot AI Tools in a Controlled Environment with Benchmarks
3. Deployment phase: Implemented with clear policies and training
4. Surveillance phaseEstablish continuous quality control and feedback mechanisms

Use the AI policy component.

• Permitted Uses: Research support, document review, initial drafting
• Limited use: Client communication, court submission (with verification)
• Prohibited useUnsupervised legal advice, autonomous decision making

Status to guarantee disclosure:

• If AI contributes substantially to legal analysis or strategy
• When client data is processed through a third-party AI platform
• When AI is used for tasks performed by traditional lawyers or paralegals
• If disclosure could affect the consideration of lawyers and clients' privileges

Real-world examples: Law firms that successfully implement the use of ethical AI may adopt a contract review system with lawyers monitoring that will check all findings and maintain client communications, whilst AI flagging potential issues. The system maintains an audit log showing which sections are AI analyzed and which sections have been human reviewed, creating accountability and transparency throughout the process.

Stay on ethical practices

Groff reiterated that core ethics rules apply to AI without exception. He also noted the need for lawyers to maintain technical capabilities, effectively supervise AI legal assistants, and engage in ongoing learning to ensure confidentiality and effective communication.

New governance frameworks, including the NIST AI risk management framework and industry-specific best practices, can help you spend time dealing with these issues as technology matures.

Up-to-date status regarding the use of ethical technologies:

• Consider courses and certifications that specifically address AI ethics in legal practices
• Join the Law Technical Committee through the Bar Association
• Sign up for Free Monthly Thomson Reuters AI Newsletter To gain insight into how AI is transforming the legal industry and converting notifications for future webinars.
• Join legal conferences to understand new capabilities

How to Choose an Ethical Legal AI Solution

Groff then put together the session by summarizing the key points discussed, reminding legal experts of their practices to ethically and effectively integrate the importance of continuous learning and adaptation. He closed it with a reminder Technology needs to be strengthened rather than replaced Human elements of law.

Groff clearly distinguished them from common AI tools It is specifically designed for legal applicationshighlighting the importance of choosing the right tool for a particular legal task.

Legal AI Solution Evaluation Criteria:

• Accuracy: Check facts and legal accuracy for reliable sources
• Confidentiality: Check the Data Retention Terms and Use Policy
• terms of service: Assessing licenses, liability, and compliance with corporate policies
• Integration: Consider compatibility with existing workflow systems
• training: Domain-specific legal training and general-purpose models

As AI technology continues to evolve, so will ethical considerations surrounding its use in legal practices. Legal experts must continue to maintain information about new best practices and regulatory developments. Thomson Reuters is committed to providing ongoing insights and solutions to navigate this changing landscape.