Technology in the Law – Ethical Implications of Artificial Intelligence in the Practice of Law

When the legal canons were established in 1908, they did not contemplate the technology being used in the practice of law today. Since that time, the practice of law has changed drastically.

First with the telephone and fax machine, then wireless phones, the internet, email, and smart phones. Mobile devices and emerging technologies are further expanding our horizons and changing the way lawyers practice law. However, with new technology comes new ethical concerns, and lawyers must constantly evaluate whether a particular technology can or should be used in his or her practice. The emergence of artificial intelligence (“AI”) is one of the newest technologies to pose ethical concerns in the practice of law. Is it ethical to use AI? Is it unethical not to use AI? How can it be used responsibly and ethically to allow lawyers to better serve their clients?

Technology is ubiquitous and has hit the legal profession, even though lawyers have not traditionally been early adopters of technology. At every step of technological development lawyers must ask the following questions: “is it professional, is it safe, does it protect confidentiality?” In the early days of technology, large law firms resisted the use of the telephone amid concerns of lack of confidentiality and professionalism. In response to these ongoing concerns, the ABA has continually issued opinions regarding emerging technology such as the cordless phone, fax machine, smart phone, and email.

Rules Governing the Use of Technology

ABA Model Rule 1.1 requires a lawyer to have the knowledge and skills necessary to competently represent the client. The lawyer is prohibited from taking on the representation if the lawyer does not have the required skills to handle the client’s matter. In such a case the lawyer must turn down the representation, spend time to acquire the necessary skills, or can associate with a competent lawyer. Comment 8 to Rule 1.1 provides that lawyers must keep abreast of relevant technology and understand the benefits and risks. In states adopting this obligation, lawyers must understand what technology is available and, if they choose to use it, they must understand how it works.

Rule 1.4 requires a lawyer to reasonably consult with the client about the means by which the client’s objectives are to be accomplished. In terms of technology, this would require the lawyer to discuss the different applications that may be used in the representation and how they might impact the client and his or her confidential information.

Rules 5.1 and 5.3 involve a lawyer’s responsibilities with respect to supervising a non-lawyer who may be hired with respect to the client’s representation. For example, if the lawyer will be uploading the client’s data to an e-discovery platform, how does that platform protect confidential information? Who has access to the data? Where is it stored? Does the platform retain any rights to the uploaded data? It is the lawyer’s obligation to ensure that the vendor’s conduct is compatible with the lawyer’s professional obligations.

Confidentiality has been a central theme throughout history with respect to emerging technologies used by lawyers. Rule 1.6 generally provides that, “A lawyer shall not reveal information relating to the representation of a client unless the client gives informed consent.” Thus, using technology that may put a client’s information at risk is understandably a significant and real concern. With respect to the use of AI and emerging technology in a law practice, a lawyer needs to be additionally concerned with how the client’s information is being stored, whether it remains on the vendor’s server, and what the vendor is doing to maintain confidentiality and protect against unauthorized exposure.

Rule 1.5 requires lawyers to charge reasonable fees. Because technology and AI are generally utilized to reduce the amount of time spent on a particular task, failing to use available technology may cause an attorney to overcharge a client. Take for example, a patent attorney who used AI to write a system for routing communications in a computer network. The system automatically prepared the specs and the application. The attorney spent two hours reviewing the draft as opposed to the twenty hours it would have taken to prepare it himself. The $2,500 preparation fee was significantly less than the attorney’s billable time. If technology can help perform a task better, faster, and cheaper, an attorney should consider whether it should be used. If the attorney chooses not to use the technology, he or she may have to consult with the client to explain why.

Cyber Security

Cyber security is a threshold topic that must be addressed in connection with emerging technologies used by lawyers. Law firms, both big and small, are facing emerging threats from hackers and cyber thieves searching for ways to obtain confidential information or gain access to the firm’s trust account. Many smaller firms have traditionally believed cyber crime was a large firm problem, and therefore did not implement strict security measures to safeguard their systems. Now outside actors are targeting these smaller firms, compromising the firm’s email account, and waiting for the perfect opportunity to strike. In addition to hacking the firm’s email, the criminal now has access to the client’s confidential information and secure documents, causing an ethical nightmare. The following is a non-exhaustive list of some initial steps law firms should take to prevent data breaches in the light of emerging technology:

• Use two factor authentication;
• Implement awareness training programs (automated phishing campaigns);
• Perform vulnerability scanning;
• Use of encryption for all devices (laptops, phones, and USB drives) in case of loss or theft;
• Make sure all devices are known, logged;
• Ensure all mobile devices are secured, locked.

When these strategies are implemented, the firm’s risk factor is significantly reduced.

What is AI?

There has recently been a wave of AI-driven technologies because of major advances in computer processing power. These developments mean that AI is poised to radically change businesses, entire bodies of knowledge and our daily lives. To note just a few recent examples in the news: AI is helping chemists in the lab by producing blueprints to create molecules, Chinese hospitals are hoping that AI can alleviate a shortage of doctors and help treat patients, researchers have trained artificial neural networks to detect signs of life on other planets, and AI is powering self-driving vehicles that are increasingly on the road with us.

AI is a category of technologies in which computers learn and take on complex tasks that previously were possible only through substantial human effort. AI is a broad term that covers several techniques that allow machines to simulate human intelligence. Natural language processing, for example, is an area of AI that is especially relevant for law firms. It has to do with how computers interact with human languages and allows users to communicate with an AI system as if speaking to another lawyer, and then receive only highly relevant answers.

Lawyers are using AI in the following common ways:

• Text extraction and classification – technology assisted review
• Natural language processing – legal research, Google
• Machine learning – due diligence review
• Neural networks, deep learning – Google translate
• Rule based systems – expense management

How is AI Used in Law?

AI is being integrated into the practice of law in a variety of ways. It starts with the ability to understand language in natural terms and extract text. But natural language capabilities are not perfect. In a 2015 study, a law professor compared six of the top legal databases and entered the identical search in the same jurisdictional database of reported cases. Only seven decisions appeared in all six results, and 40% of the decisions only appeared in one result. The take away – different algorithms lead to different results. If the starting point for research is a keyword search, the divergent results in each of these six databases may frame the research in a very different way. Lawyers must continually evaluate whether the tool being used is competent to perform the assigned task, and whether the results are competent.

Many lawyers are turning to virtual digital assistants such as Siri, Alexa, Cortana, or Google Assistant for help scheduling appointments, getting answers to questions, conducting research, billing, and other mundane tasks. While these services can help lawyers with small tasks, there are some confidentiality issues to consider. When a digital assistant is used, everything that is being said gets sent over the internet to the digital assistant host company for processing, where the user’s speech is often analyzed and stored in order that the host company can improve its digital assistant. Most digital assistants require a trigger or “wake” word that fires them up to be able to respond to a question or command. That means the microphone must remain on in order to listen for the trigger word, potentially listening in on everything a lawyer is saying. So, if an attorney has just finished a client meeting and asks Siri to “record 0.3 hours on Bill Smith matter for conference with clients regarding XYZ matter,” Siri or other virtual assistant may also store part of the prior and subsequent conversation, leading to confidentiality concerns.

The development of AI has also lead to other functions. By examining a particular judge’s decisions, AI can give feedback on which judge may be favorable to a case, predict whether a certain case will win or lose at trial, provide guidance on how to draft an argument for a particular judge, or even prepare a pleading based on success in other cases. However, whether the results of the AI program are reliable is another matter entirely. If the lawyer blindly relies on AI, he or she will be the one exposed for potential liability for relying on technology without performing sufficient due diligence. On the other hand, could it be considered unethical not to use these tools for the benefit of the client?

Is AI Reliable?

Before an attorney uses AI as a tool, he or she needs to evaluate the security of the tool and how it could impact the client. At the end of the day, the client will not care whether the attorney was breached or the vendor – the attorney may still be held responsible. Thus, an attorney should investigate what kind of security the vendor uses. Where does data live, who has access, and what does the vendor do with it? Is the data being uploaded being used for the client’s interest or does the vendor have rights to it? Was the algorithm being used created to benefit the client?

These questions highlight an interesting issue pertaining to AI – the fact that AI can lead to biased decision making. How the AI was trained can lead to varied results. An attorney should look into what data was used to train the AI. Data sets that are the product of human decisions may reflect the preferences of trainer. If AI is trained on a biased set of data, the AI will learn and propagate those biases. Additional questions to address this issue include determining who trained the AI, whether the training data was trustworthy, and could the data reflect bias toward the other party’s position?

Terms of Service

Before using AI in a law practice, an attorney should carefully review the vendor’s terms of service. Terms vary greatly among vendors, and can lead to questions regarding disclosure of confidential information. Service terms are critical for number of reasons. They govern who has a right to the data output, or whether the client only has a license to the data. Some vendors retain the right to the derivative of the information provided. For example, Microsoft Office 365’s terms give the vendor extremely broad rights in uploaded data. To avoid inadvertent disclosure, the attorney’s only recourse is to not use the tool.

Another important aspect to consider with respect to the terms of service is the exit strategy. What does the vendor do with the data after the contract is terminated, or after the vendor goes out of business? How long does the vendor retain the data once the contract is concluded? Is any part of the uploaded data retained? For example, some vendors retain obscured data in order to further train the AI algorithm.

The Future of Technology in the Law

There’s no question that the entire legal profession will eventually be transformed by AI and emerging technology. Less time will be spent on repetitive tasks, freeing attorneys to perform other more creative and efficient work. Lawyers are ethically obligated to learn about new technology and to understand its potential impact on the practice of law. Rejecting technology and refusing to understand it could leave an attorney open to discipline.