|
YueYuFirstPaper 4 - 10 May 2025 - Main.YueYu
|
| |
| META TOPICPARENT | name="FirstPaper" |
|---|
The Cost of Convenience: How AI Threatens the Right to Effective Assistance of Counsel | |
<
< | -- By YueYu - 19 Mar 2025 | >
> | -- By YueYu - 19 Mar 2025 (10 May 2025 Second Revision) | | | “Freedom and love are dear to me,
My life I give, sweet love, for thee, | | | By Petőfi Sándor (translated by William N. Loew)
| |
<
< | Liberty encompasses not only physical freedom but also the freedom to think. Yet today, our thinking is increasing restrained within the framework set by the so-called “Artificial Intelligence”. While our minds should remain an open field, we now may lose sight of all the possibilities beyond the tracks AI presents with us. We give liberty, not for life or love, but simply for “convenience”. | >
> | Liberty encompasses not only physical freedom but also the freedom to think. Yet today, our thinking is increasingly restrained by so-called “Artificial Intelligence”. While our minds should remain an open field, we now lose sight of possibilities beyond the tracks AI presents with us. We give liberty, not for life or love, but simply for “convenience”. | | |
The Right to Effective Assistance of Counsel is Compromised When Lawyers Rely Blindly on AI.
| |
<
< | It is convenient for lawyers to use AI for legal research, drafting documents, and even selecting jurors. Proponents of AI claim that using it can greatly improve efficiency, allowing lawyers to focus on tasks that require human intelligence, such as strategic thinking. This may be true, but when lawyers began to ask AI for answers to every question and blindly rely on the information and conclusions AI provides, they risk losing the ability to think. | >
> | AI offers lawyers convenience in legal research, document drafting, and even jury selection. Proponents claim that it greatly improves efficiency, enabling lawyers to focus on tasks that require human intelligence, such as strategic thinking. This may be true, but when lawyers begin blindly relying AI for every answer, they risk losing the ability to think. | | | | |
<
< | Sixth Amendment guarantees the accused in criminal prosecutions the right to have the assistance of counsel. The Court has held that the right to counsel is the right to the effective assistance of counsel and it can be deprived simply by counsel’s failure to render adequate legal assistance. McMann v. Richardson; Strickland v. Washington | >
> | The Sixth Amendment guarantees the accused in criminal prosecutions the right to have the assistance of counsel, construed as the right to the effective assistence of counsel. McMann v. Richardson The standard set by Strickland v. Washington requires the accused to prove (1) the defense attorney was objectively deficient, and (2) a reasonable probability that a competent attorney would have led to a different outcome. It’s a high bar, as courts often defer to “strategic choices”, even foolish ones, by defense counsel. As a result, most such claims fail. Ineffective assistance of counsel may cause more severe damages in criminal cases where an individual’s liberty or even life is at stake. That said, legal profession advocates right to counsel in civil cases, and some states have created limited statutory rights in civil cases. | | | | |
<
< | A lawyer who blindly relies on AI cannot provide adequate legal assistance. Information provided by AI can be wrong, fabricated or misleading. We’ve seen cases where lawyers use fabricated cases to support claims or arguments. For example, in 2023, two lawyers were sanctioned by the Southern District Court of New York for quoting six non-existent cases generated by ChatGPT in a legal brief. One of them testified, “I just was not thinking that the case could be fabricated, so I was not looking at it from that point of view.” While the judge found the lawyers acted in bad faith, I tend to believe that they just relied blindly on ChatGPT? without thinking. | >
> | Unlike mistakes that may be interpreted as “strategies”, such as a counsel sleeping through parts of a trial (McFarland v. State), blindly citing AI-generated, fictitious cases or laws cannot be excused as a “strategy”. A lawyer’s failure to find supporting authority due to blind reliance on AI strengthens a claim that competent representation could have produced a different outcome.
A more fundamental threat is that, lawyers and law students are gradually losing motivation and ability to read, write and think. In a recent civil case, two lawyers were sanctioned by the Southern District Court of New York for quoting six non-existent cases generated by ChatGPT in a legal brief. One testified, “I just was not thinking that the case could be fabricated, so I was not looking at it from that point of view.” While the judge found the lawyers acted in bad faith, I tend to believe that they just relied blindly on ChatGPT? without thinking. Such lawyers cannot provide adequate legal assistance, but under the prevailing standard, it’s nearly impossible to claim ineffective assistance of counsel solely based on a lawyer’s lack critical thinking.
Compared to the constitutional remedy, professional discipline provides no direct remedy for the accused. Competent representation required by ethical rules, such as Rule 1.1 of the Model Rules of Professional Conduct, is a lower bar than the constitutional standard since it doesn’t require showing a “different outcome”. However, like the constitutional remedy, bar discipline may address only specific instances of incompetency, but is difficult to apply to the situation described immediately above. | | |
The Right to Effective Assistance of Counsel is Undermined When Defendants Are Represented by a “Robot Lawyer”.
| |
<
< | It’s convenient for criminal defendants turn to an “AI Lawyer” for legal advice, drafting legal documents, or even being “represented” by a robot lawyer. But does this meet the constitutional requirements? The Supreme Court has never ruled on how the Sixth Amendment as applied to non-human counsel. However, if we interpret the right to counsel from its plain meaning, it should be understood as the right to “human counsel”. A defendant choosing a robot lawyer could be seen as waiving the right to counsel. Such waiver, however, should be made knowingly, voluntarily and intelligently. | >
> | It’s convenient for criminal defendants turn to an “AI Lawyer” for legal advice, legal document drafting, or even representation. But does this meet the constitutional requirements? The Supreme Court has never ruled on how the Sixth Amendment as applied to non-human counsel. However, a plain meaning interpretation implies a right to “human counsel”. Choosing a robot lawyer could be seen as a waiver of that right. Such waiver, however, should be made knowingly, voluntarily and intelligently. | | | | |
<
< | Many AI-based legal service providers promote that “AI lawyers” are just as competent as, or even more competent than, human lawyers due to their high efficiency and accuracy. In a complaint filed by the Federal Trade Commission (FTC) against DoNotPay, a company that promoted its online subscription service as “the world’s first robot lawyer”, the FTC asserted that the so-called robot lawyer failed to live up to claims that it was an adequate substitute for the expertise of a human lawyer. If a defendant waives the right to counsel based on such misleading claims, it is hard to say that such waiver was made intelligently. | >
> | AI-based legal service providers promote that “AI lawyers” are just as competent as, or even more competent than, human lawyers due to high efficiency and accuracy. In a Federal Trade Commission (FTC) complaint against DoNotPay, which promoted its online subscription service as “the world’s first robot lawyer”, the FTC asserted that the so-called robot lawyer failed to live up to claims that it was an adequate substitute for the expertise of a human lawyer. If a defendant waives the right to counsel based on such misleading claims, it is doubtful that such waiver was made intelligently. Therefore, different from other forms of incompetency, an accused using a robot lawyer may not even realize that their rights have been violated. | | | | |
<
< | There are other issues as well. When an accused is “represented” by a robot lawyer, attorney-client privilege may not apply to the information shared with the robot. This could allow the government to compel the robot lawyer's service provider to disclose confidential information. | >
> | There are other issues. When an accused is “represented” by a robot lawyer, attorney-client privilege may not apply to the information shared with the robot, and providers may be compelled to disclose that data to the government. | | |
Solutions
Stop Relying Blindly on AI
| |
<
< | I am not a fan of the Russian proverb “Trust, but verify”. If something needs to be verified, it can’t be trusted. AI is merely a tool that assists lawyers in collecting and analyzing data, and it is lawyers’ responsibility to verify information generated by AI. In addition, we need to stop asking AI questions without first considering them by ourselves. It starts with each of us because we should not expect others to safeguard our freedom to think. | >
> | I am not a fan of the Russian proverb “Trust, but verify”. If something needs verification, it can’t be trusted. AI is merely a tool assisting lawyers in collecting and analyzing data, and it is lawyers’ responsibility to verify AI-generated information. In addition, we must stop relying on AI-generated answers without first considering them by ourselves. It starts with each of us because we should not expect others to safeguard our freedom to think. | | |
Regulating Service Providers
| |
<
< | In February 2025, FTC finalized an order requiring DoNotPay? to stop making deceptive claims about the abilities of its “robot lawyer”. A disclaimer on the company’s website states, “DoNotPay is not a law firm, is not licensed to practice law, and is not equivalent to the services of a licensed lawyer.” While it’s unclear if this disclaimer was added due to the FTC order, its effectiveness is diminished by its small font and placement among other content. Aside from DoNotPay? , there are numerous other AI legal service providers on the market. These companies should be regulated to prevent false or misleading advertising, ensuring that defendants can make informed decisions about whether to waive their right to human counsel. | >
> | In February 2025, FTC ordered DoNotPay? to stop making deceptive claims about its “robot lawyer”. The company’s website now includes a disclaimer: “DoNotPay is not a law firm, is not licensed to practice law, and is not equivalent to the services of a licensed lawyer.” It’s unclear if this was added in response to the FTC order, but even so, its effectiveness is diminished by its small font and placement among other content. Companies like DoNotPay? should be regulated to prevent false or misleading advertising, ensuring that defendants can make informed decisions about whether to waive their right to human counsel. | | | | |
<
< |
It seems to me that there is less to this draft than meets the eye. The basic proposition is that incompetence vitiates the right to effective assistance. But you do not actually show the standards applicable, which would somewhat undercut the breadth of the argument you are making. And you do not show why one form of bad legal research or incompetent writing is either more unconstitutional or more likely than all the other causes of bad legal practice that either do or do not rise to the level of a constitutional failure when they happen to occur in criminal as opposed to non-criminal proceedings.
Would it not have made some sense to consider the role of bar discipline as opposed to Sixth Amendment relief against criminal conviction? Your only significant example concerned the submission of civil litigation paper containing unchecked citations, which both happens all the time and is handled in the case you show by sanctions. At what point should bar discipline be appropriate?
| | |
You are entitled to restrict access to your paper if you want to. But we all derive immense benefit from reading one another's work, and I hope you won't feel the need unless the subject matter is personal and its disclosure would be harmful or undesirable.
To restrict access to your paper simply delete the "#" character on the next two lines: |
|
|
YueYuFirstPaper 3 - 04 May 2025 - Main.EbenMoglen
|
| |
| META TOPICPARENT | name="FirstPaper" |
|---|
| | | | |
>
> |
It seems to me that there is less to this draft than meets the eye. The basic proposition is that incompetence vitiates the right to effective assistance. But you do not actually show the standards applicable, which would somewhat undercut the breadth of the argument you are making. And you do not show why one form of bad legal research or incompetent writing is either more unconstitutional or more likely than all the other causes of bad legal practice that either do or do not rise to the level of a constitutional failure when they happen to occur in criminal as opposed to non-criminal proceedings.
Would it not have made some sense to consider the role of bar discipline as opposed to Sixth Amendment relief against criminal conviction? Your only significant example concerned the submission of civil litigation paper containing unchecked citations, which both happens all the time and is handled in the case you show by sanctions. At what point should bar discipline be appropriate?
| | |
You are entitled to restrict access to your paper if you want to. But we all derive immense benefit from reading one another's work, and I hope you won't feel the need unless the subject matter is personal and its disclosure would be harmful or undesirable.
To restrict access to your paper simply delete the "#" character on the next two lines: |
|
|
YueYuFirstPaper 2 - 25 Mar 2025 - Main.YueYu
|
|
<
< | Revision 1 is unreadable | >
> |
| META TOPICPARENT | name="FirstPaper" |
|---|
The Cost of Convenience: How AI Threatens the Right to Effective Assistance of Counsel
-- By YueYu - 19 Mar 2025
“Freedom and love are dear to me,
My life I give, sweet love, for thee,
Yet love I give for liberty.”
By Petőfi Sándor (translated by William N. Loew)
Liberty encompasses not only physical freedom but also the freedom to think. Yet today, our thinking is increasing restrained within the framework set by the so-called “Artificial Intelligence”. While our minds should remain an open field, we now may lose sight of all the possibilities beyond the tracks AI presents with us. We give liberty, not for life or love, but simply for “convenience”.
The Right to Effective Assistance of Counsel is Compromised When Lawyers Rely Blindly on AI.
It is convenient for lawyers to use AI for legal research, drafting documents, and even selecting jurors. Proponents of AI claim that using it can greatly improve efficiency, allowing lawyers to focus on tasks that require human intelligence, such as strategic thinking. This may be true, but when lawyers began to ask AI for answers to every question and blindly rely on the information and conclusions AI provides, they risk losing the ability to think.
Sixth Amendment guarantees the accused in criminal prosecutions the right to have the assistance of counsel. The Court has held that the right to counsel is the right to the effective assistance of counsel and it can be deprived simply by counsel’s failure to render adequate legal assistance. McMann v. Richardson; Strickland v. Washington
A lawyer who blindly relies on AI cannot provide adequate legal assistance. Information provided by AI can be wrong, fabricated or misleading. We’ve seen cases where lawyers use fabricated cases to support claims or arguments. For example, in 2023, two lawyers were sanctioned by the Southern District Court of New York for quoting six non-existent cases generated by ChatGPT in a legal brief. One of them testified, “I just was not thinking that the case could be fabricated, so I was not looking at it from that point of view.” While the judge found the lawyers acted in bad faith, I tend to believe that they just relied blindly on ChatGPT? without thinking.
The Right to Effective Assistance of Counsel is Undermined When Defendants Are Represented by a “Robot Lawyer”.
It’s convenient for criminal defendants turn to an “AI Lawyer” for legal advice, drafting legal documents, or even being “represented” by a robot lawyer. But does this meet the constitutional requirements? The Supreme Court has never ruled on how the Sixth Amendment as applied to non-human counsel. However, if we interpret the right to counsel from its plain meaning, it should be understood as the right to “human counsel”. A defendant choosing a robot lawyer could be seen as waiving the right to counsel. Such waiver, however, should be made knowingly, voluntarily and intelligently.
Many AI-based legal service providers promote that “AI lawyers” are just as competent as, or even more competent than, human lawyers due to their high efficiency and accuracy. In a complaint filed by the Federal Trade Commission (FTC) against DoNotPay, a company that promoted its online subscription service as “the world’s first robot lawyer”, the FTC asserted that the so-called robot lawyer failed to live up to claims that it was an adequate substitute for the expertise of a human lawyer. If a defendant waives the right to counsel based on such misleading claims, it is hard to say that such waiver was made intelligently.
There are other issues as well. When an accused is “represented” by a robot lawyer, attorney-client privilege may not apply to the information shared with the robot. This could allow the government to compel the robot lawyer's service provider to disclose confidential information.
Solutions
Stop Relying Blindly on AI
I am not a fan of the Russian proverb “Trust, but verify”. If something needs to be verified, it can’t be trusted. AI is merely a tool that assists lawyers in collecting and analyzing data, and it is lawyers’ responsibility to verify information generated by AI. In addition, we need to stop asking AI questions without first considering them by ourselves. It starts with each of us because we should not expect others to safeguard our freedom to think.
Regulating Service Providers
In February 2025, FTC finalized an order requiring DoNotPay? to stop making deceptive claims about the abilities of its “robot lawyer”. A disclaimer on the company’s website states, “DoNotPay is not a law firm, is not licensed to practice law, and is not equivalent to the services of a licensed lawyer.” While it’s unclear if this disclaimer was added due to the FTC order, its effectiveness is diminished by its small font and placement among other content. Aside from DoNotPay? , there are numerous other AI legal service providers on the market. These companies should be regulated to prevent false or misleading advertising, ensuring that defendants can make informed decisions about whether to waive their right to human counsel.
You are entitled to restrict access to your paper if you want to. But we all derive immense benefit from reading one another's work, and I hope you won't feel the need unless the subject matter is personal and its disclosure would be harmful or undesirable.
To restrict access to your paper simply delete the "#" character on the next two lines:
Note: TWiki has strict formatting rules for preference declarations. Make sure you preserve the three spaces, asterisk, and extra space at the beginning of these lines. If you wish to give access to any other users simply add them to the comma separated ALLOWTOPICVIEW list. |
|
| Revision 4 | r4 - 10 May 2025 - 14:13:13 - YueYu |
| Revision 3 | r3 - 04 May 2025 - 13:39:15 - EbenMoglen |
| Revision 2 | r2 - 25 Mar 2025 - 21:29:46 - YueYu |
| Revision 1 | r1 - 19 Mar 2025 - 17:28:09 - YueYu |
|
|
|
This site is powered by the TWiki collaboration platform.
All material on this collaboration platform is the property of the contributing authors.
All material marked as authored by Eben Moglen is available under the license terms CC-BY-SA version 4.
|
|