0:00One of the first risks that I want to talk about is bias. And it’s more pervasive than many people realize. The AI tools, they learn from data. And if that data has historical biases, you’re going to get skewed results. That’s just the way it works.

0:14We’re talking about bias sentencing predictions that unfairly target certain groups, hiring algorithms that discriminate based on zip codes, and risk assessment tools that perpetuate historical inequities.

0:29A 2023 study by the MIT Technology Review found that some AI tools trained on historical court data showed significant bias against defendants from certain de demographic groups. Now that’s not just a technological glitch. It’s a serious civil rights concern.

0:48Then there’s accuracy or sometimes the complete lack of accuracy because AI can generate fast u answers very quickly but sometimes it’s creating information that simply isn’t true and we call these hallucinations or instances where AI confidently presents false information.

1:05You might get a case citation that doesn’t exist, contract terms that are legally problematic, or legal analysis that’s analysis that sounds sophisticated but is actually fundamentally flawed.

1:24In 2023, lawyers in New York got sanctioned for submitting a brief containing six fict fictitious case citations generated by Chat GPT. Now, the issue came to light when opposing council and the court could not locate the cited cases.

1:39This led to sanctions for the lawyers failure to verify the accuracy of their findings and their subsequent misleading statements to the court.

1:48Now, importantly, the judge noted there’s nothing inherently improper about using a AI tools for legal assistance, but attorneys have a gatekeeping role to ensure the accuracy of their findings under federal rule of civil procedure 11B.

2:02The lawyers failure to disclose the use of chat GPT initially and their persistence in defending the fake fake citations after concerns were raised constituted bad faith in the court’s view.

2:18The attorneys then admitted to using chatbt and believed that it was a reliable search tool and expressed regret for not verifying the citations. They just assumed they were unpublished or inaccessible cases.

2:31The sanctions included a $5,000 fine and an order to notify the judges uh that were falsely identified as authors of the fake cases.

2:42Client data security is another major challenge with AI because when you input sensitive information into cloud-based AI tools that aren’t properly secured, you’re risking data breaches that it could expose everything from merger details to personal information in a family law case, for example.

2:59That’s a fast path to e ethical violations, malpractice claims, and obviously severely damaged client relationships and your firm’s reputation.

3:12So, there’s an over reliance trap that’s happening that’s catching more lawyers than you might expect because when you start to depend too heavily on AI, you can forget that you’re the lawyer and the algorithm is not.

3:25It’s easy to trust the machine more than your own professional judgment. And that’s when serious mistakes start to happen. AI might miss the nuances that an attorney would more easily catch, can misinterpret context, or provide technically correct but practically inappropriate advice.

3:46It’s more AI is more black and white than the gray areas that allow us as lawyers to push for advocacy and justice in any given situation.

3:56There’s also the evolving question of transparency and professional responsibility. For example, if you’re using AI to draft motions, research cases, or analyze contracts, what are your disclosure obligations? Do you need to tell the client, the court, opposing council?

4:16These rules are still developing, but the trend is clearly towards more transparency, not less.

4:23Professional competence also takes on a new meaning in the AI era. So, you need to understand not just the law, but also the capabilities and the limitations of the tools you’re using.

4:34Now, that means staying current with AI developments, understanding how these systems work, and knowing when human judgment is irreplaceable when you’re using AI.

4:44So, these ris risks are serious and they’re real, and they’re setting the stage for why we need to talk about our professional responsibilities in the first place.

4:52The opportunities are tremendous. AI can make us more efficient, more accurate, and more accessible to clients. But those benefits come with ethical obligations and potential pitfalls that can seriously damage your practice if you’re not careful.

5:07So the key is approaching AI with enthusiasm and great caution. We need to embrace these benefits but also at the same time build safeguards against the risks.

5:19And that means understanding the professional rules, implementing strong verification procedures, and maintaining confidentiality while never losing sight of our fundamental obligation to provide competent representation.