{"id":166760,"date":"2026-11-04T15:50:36","date_gmt":"2026-11-04T23:50:36","guid":{"rendered":"https:\/\/xira.com\/p\/2026\/11\/04\/ai-connectors-for-e-discovery-great-to-drive-but-there-are-speed-limits\/"},"modified":"2026-11-04T15:50:36","modified_gmt":"2026-11-04T23:50:36","slug":"ai-connectors-for-e-discovery-great-to-drive-but-there-are-speed-limits","status":"publish","type":"post","link":"https:\/\/xira.com\/p\/2026\/11\/04\/ai-connectors-for-e-discovery-great-to-drive-but-there-are-speed-limits\/","title":{"rendered":"AI Connectors For E-Discovery: Great To Drive, But There Are Speed Limits"},"content":{"rendered":"
\"\"
<\/figcaption><\/figure>\n

A recent article<\/a> reminded me again of the dangers of unfettered use of AI tools, particularly by business employees. Employees who mistakenly believe that what they put in and get out is private. It\u2019s not.<\/p>\n

According to the article, Reveal<\/a> announced that its Onna platform can now collect conversations and other materials that employees generate in Claude Enterprise. Reveal provides AI platforms for e-discovery purposes; among other things, Onna already provided similar capabilities for users of ChatGPT and Gemini. And other e-discovery providers offer similar tools. This means that the prompts employees write and what those prompts generate can now be easily searched and collected for review.\u00a0<\/p>\n

The idea, of course, is that it will help businesses and their lawyers respond to discovery requests more efficiently and accurately. Or be able to easily obtain and review this material for internal investigations.<\/p>\n

A Wake-Up Call<\/strong><\/p>\n

No doubt this is all true and provides a good tool. But it\u2019s also a good reminder for in-house legal departments. I fear that lots of employees are using Claude and other AI tools for all sorts of things, thinking that what they put in there is, well, private.\u00a0<\/p>\n

For legal and discovery purposes, this lack of understanding could be disastrous. If employees don\u2019t understand that whatever they put in Claude could be discoverable, they could be creating all sorts of \u201csmoking guns.\u201d Documents that, at the very least, could be hard to explain.\u00a0<\/p>\n

Of course, that was true before Onna came along, and its integration doesn\u2019t make that risk any greater than it always was. But it\u2019s an important reminder that all the AI stuff that goes in and comes out may be discoverable. Which your employees may not appreciate.\u00a0<\/p>\n

It underscores your responsibility for employee education and training, as I have discussed before. It\u2019s like the early days of email. I saw too many instances of hard-to-explain emails coming to light in litigation. All because employees didn\u2019t realize that their emails were discoverable. Or fully appreciate that fact. Employees indeed need to understand litigation, its dangers and costs. And the parameters of discovery.\u00a0<\/p>\n

The \u2018Private\u2019 File<\/strong><\/p>\n

I had a case one time where we spent considerable time pulling together the documents and materials for production. I just happened to ask one key employee at the tail end of the process if we had everything. He sort of shrugged and said he guessed so. Then he added, \u201cYou got everything but the private file I keep on my home computer.\u201d Crap. Of course, that \u201cprivate file\u201d contained all sorts of notes and data that were highly relevant to the case.<\/p>\n

The point is that in-house and outside counsel can\u2019t assume that employees know the risks without some education. Without some time being spent in training.\u00a0<\/p>\n

And believe me, that\u2019s time well spent. You reduce the risk of bad documents being inadvertently created that could adversely affect the outcome. And think of this: you will also reduce your legal spend. Take it from me, your lawyers will spend countless hours dealing with documents that should never have been created. Or at least never worded the way they were.<\/p>\n

An Exponential Danger<\/strong><\/p>\n

And this danger grows exponentially when you start thinking about the use of AI. It\u2019s so easy and tempting to use for anything and everything. You often don\u2019t think twice about what you put in it. And because it\u2019s so conversational, it\u2019s easy to add things and play around with prompts in ways that could be hard to explain later. It\u2019s easy to forget you aren\u2019t talking to a human. It\u2019s easy to forget you\u2019re making a record.\u00a0<\/p>\n

Sure, you may be able to later argue that some of this stuff might, and I emphasize might, be protected by the attorney-client or work product privileges. But the law here is unsettled.<\/p>\n

And the work product privilege is not absolute in any event. To be honest, I would much rather not have documents around that I have to hope are covered by it.\u00a0<\/p>\n

It\u2019s also a good reminder that what employees put into AI using their own devices can also be discoverable. Their cell phones. Their home laptops. There are no private files when it comes to litigation.<\/p>\n

It Belongs To The Company<\/strong><\/p>\n

There\u2019s another issue here that I almost hate to bring up. Generally, anything an employee puts into a work platform belongs to the company. And the company is usually entitled to access that material any time it wants. I know that. In-house legal knows that. But do your employees all know that?\u00a0<\/p>\n

Education about AI and the lack of privacy is a good time to reinforce the point that employers have the right to see what employees put in and take out of the employer\u2019s AI systems. It might save heartburn all the way around if this were specifically covered when you talk to them about litigation risks.<\/p>\n

Training Need Not Be Hard<\/strong><\/p>\n

And the training and education need not be all that difficult. I used to give a couple of sessions a year to good clients and their employees about how not to create bad documents but still do their job. I would throw in a few examples from real well-known cases that more than made the point. For in-house counsel, I can tell you your outside lawyers would be more than happy to do similar sessions at no charge. If they\u2019re worth their salt and business savvy, that is.<\/p>\n

Obey The Speed Limit<\/strong><\/p>\n

So it\u2019s great that companies like Reveal come up with tools that make the e-discovery process more efficient and less time-consuming. It would be even better if the searches these tools enable didn\u2019t turn up prejudicial and embarrassing materials. That would be a win-win.\u00a0<\/p>\n

But if you want that win-win, you\u2019ve got to take time to educate your workforce about the shiny new AI tool. A tool that\u2019s great for them to drive, but you sure don\u2019t want them going 100 mph. In a school zone.<\/p>\n

\n

Stephen Embry is a lawyer, speaker, blogger, and writer. He publishes\u00a0<\/em><\/strong>TechLaw Crossroads<\/em><\/strong><\/a>, a blog devoted to the examination of the tension between technology, the law, and the practice of law<\/em><\/strong>.\u00a0<\/p>\n

\n

The post AI Connectors For E-Discovery: Great To Drive, But There Are Speed Limits<\/a> appeared first on Above the Law<\/a>.<\/p>\n

\"\"
<\/figcaption><\/figure>\n

A recent article<\/a> reminded me again of the dangers of unfettered use of AI tools, particularly by business employees. Employees who mistakenly believe that what they put in and get out is private. It\u2019s not.<\/p>\n

According to the article, Reveal<\/a> announced that its Onna platform can now collect conversations and other materials that employees generate in Claude Enterprise. Reveal provides AI platforms for e-discovery purposes; among other things, Onna already provided similar capabilities for users of ChatGPT and Gemini. And other e-discovery providers offer similar tools. This means that the prompts employees write and what those prompts generate can now be easily searched and collected for review.\u00a0<\/p>\n

The idea, of course, is that it will help businesses and their lawyers respond to discovery requests more efficiently and accurately. Or be able to easily obtain and review this material for internal investigations.<\/p>\n

A Wake-Up Call<\/strong><\/p>\n

No doubt this is all true and provides a good tool. But it\u2019s also a good reminder for in-house legal departments. I fear that lots of employees are using Claude and other AI tools for all sorts of things, thinking that what they put in there is, well, private.\u00a0<\/p>\n

For legal and discovery purposes, this lack of understanding could be disastrous. If employees don\u2019t understand that whatever they put in Claude could be discoverable, they could be creating all sorts of \u201csmoking guns.\u201d Documents that, at the very least, could be hard to explain.\u00a0<\/p>\n

Of course, that was true before Onna came along, and its integration doesn\u2019t make that risk any greater than it always was. But it\u2019s an important reminder that all the AI stuff that goes in and comes out may be discoverable. Which your employees may not appreciate.\u00a0<\/p>\n

It underscores your responsibility for employee education and training, as I have discussed before. It\u2019s like the early days of email. I saw too many instances of hard-to-explain emails coming to light in litigation. All because employees didn\u2019t realize that their emails were discoverable. Or fully appreciate that fact. Employees indeed need to understand litigation, its dangers and costs. And the parameters of discovery.\u00a0<\/p>\n

The \u2018Private\u2019 File<\/strong><\/p>\n

I had a case one time where we spent considerable time pulling together the documents and materials for production. I just happened to ask one key employee at the tail end of the process if we had everything. He sort of shrugged and said he guessed so. Then he added, \u201cYou got everything but the private file I keep on my home computer.\u201d Crap. Of course, that \u201cprivate file\u201d contained all sorts of notes and data that were highly relevant to the case.<\/p>\n

The point is that in-house and outside counsel can\u2019t assume that employees know the risks without some education. Without some time being spent in training.\u00a0<\/p>\n

And believe me, that\u2019s time well spent. You reduce the risk of bad documents being inadvertently created that could adversely affect the outcome. And think of this: you will also reduce your legal spend. Take it from me, your lawyers will spend countless hours dealing with documents that should never have been created. Or at least never worded the way they were.<\/p>\n

An Exponential Danger<\/strong><\/p>\n

And this danger grows exponentially when you start thinking about the use of AI. It\u2019s so easy and tempting to use for anything and everything. You often don\u2019t think twice about what you put in it. And because it\u2019s so conversational, it\u2019s easy to add things and play around with prompts in ways that could be hard to explain later. It\u2019s easy to forget you aren\u2019t talking to a human. It\u2019s easy to forget you\u2019re making a record.\u00a0<\/p>\n

Sure, you may be able to later argue that some of this stuff might, and I emphasize might, be protected by the attorney-client or work product privileges. But the law here is unsettled.<\/p>\n

And the work product privilege is not absolute in any event. To be honest, I would much rather not have documents around that I have to hope are covered by it.\u00a0<\/p>\n

It\u2019s also a good reminder that what employees put into AI using their own devices can also be discoverable. Their cell phones. Their home laptops. There are no private files when it comes to litigation.<\/p>\n

It Belongs To The Company<\/strong><\/p>\n

There\u2019s another issue here that I almost hate to bring up. Generally, anything an employee puts into a work platform belongs to the company. And the company is usually entitled to access that material any time it wants. I know that. In-house legal knows that. But do your employees all know that?\u00a0<\/p>\n

Education about AI and the lack of privacy is a good time to reinforce the point that employers have the right to see what employees put in and take out of the employer\u2019s AI systems. It might save heartburn all the way around if this were specifically covered when you talk to them about litigation risks.<\/p>\n

Training Need Not Be Hard<\/strong><\/p>\n

And the training and education need not be all that difficult. I used to give a couple of sessions a year to good clients and their employees about how not to create bad documents but still do their job. I would throw in a few examples from real well-known cases that more than made the point. For in-house counsel, I can tell you your outside lawyers would be more than happy to do similar sessions at no charge. If they\u2019re worth their salt and business savvy, that is.<\/p>\n

Obey The Speed Limit<\/strong><\/p>\n

So it\u2019s great that companies like Reveal come up with tools that make the e-discovery process more efficient and less time-consuming. It would be even better if the searches these tools enable didn\u2019t turn up prejudicial and embarrassing materials. That would be a win-win.\u00a0<\/p>\n

But if you want that win-win, you\u2019ve got to take time to educate your workforce about the shiny new AI tool. A tool that\u2019s great for them to drive, but you sure don\u2019t want them going 100 mph. In a school zone.<\/p>\n

\n

Stephen Embry is a lawyer, speaker, blogger, and writer. He publishes\u00a0<\/em><\/strong>TechLaw Crossroads<\/em><\/strong><\/a>, a blog devoted to the examination of the tension between technology, the law, and the practice of law<\/em><\/strong>.\u00a0<\/p>\n

\n

The post AI Connectors For E-Discovery: Great To Drive, But There Are Speed Limits<\/a> appeared first on Above the Law<\/a>.<\/p>\n","protected":false},"excerpt":{"rendered":"

A recent article reminded me again of the dangers of unfettered use of AI tools, particularly by business employees. Employees who mistakenly believe that what they put in and get out is private. It\u2019s not. According to the article, Reveal announced that its Onna platform can now collect conversations and other materials that employees generate […]<\/p>\n","protected":false},"author":3,"featured_media":166761,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"_et_pb_use_builder":"","_et_pb_old_content":"","_et_gb_content_width":"","footnotes":""},"categories":[16],"tags":[],"class_list":["post-166760","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-above_the_law"],"_links":{"self":[{"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/posts\/166760","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/comments?post=166760"}],"version-history":[{"count":0,"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/posts\/166760\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/media\/166761"}],"wp:attachment":[{"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/media?parent=166760"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/categories?post=166760"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/tags?post=166760"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}