{"id":162224,"date":"2026-09-03T12:06:28","date_gmt":"2026-09-03T20:06:28","guid":{"rendered":"https:\/\/xira.com\/p\/2026\/09\/03\/from-associate-to-doppelganger-the-evolution-of-the-legal-ai-persona\/"},"modified":"2026-09-03T12:06:28","modified_gmt":"2026-09-03T20:06:28","slug":"from-associate-to-doppelganger-the-evolution-of-the-legal-ai-persona","status":"publish","type":"post","link":"https:\/\/xira.com\/p\/2026\/09\/03\/from-associate-to-doppelganger-the-evolution-of-the-legal-ai-persona\/","title":{"rendered":"From \u2018Associate\u2019 To \u2018Doppelganger\u2019: The Evolution Of The Legal AI Persona"},"content":{"rendered":"
\"\"
\u201cAI as a Teammate\u201d panelists, from left, Jeff Cox of Steno, Damien Riehl of Clio, Daniel Himmel of Legora, and Nick Hafen of BYU Law School.
<\/figcaption><\/figure>\n

If there\u2019s one overarching takeaway from ILTACON 2026, it\u2019s that legal tech has hit the big time<\/a>.<\/p>\n

As Above the Law\u2019s Stephen Embry noted, this year\u2019s conference was CES-like<\/a>, taking over the 700,000-square-foot Gaylord Opryland convention space in Nashville and drawing over 5,700 attendees.\u00a0<\/p>\n

Along with the sheer scale of ILTACON, the event also revealed the industry\u2019s rapid pace of change \u2014 even just over the past few months.\u00a0<\/p>\n

Take a pair of panels featuring Steno, the legal tech company focused on court reporting and litigation support.<\/p>\n

Back in March, Steno was featured in a panel at Legalweek, and one participant \u2014 a technology professional at a major law firm \u2014 was warning attendees of a big problem<\/a> on the horizon.\u00a0<\/p>\n

The problem, he said, was training junior lawyers, and he was worried the industry could be like \u201ca snake eating its own tail.\u201d\u00a0<\/p>\n

That\u2019s because of an apparent paradox: Today\u2019s senior lawyers can vet AI outputs and apply advanced judgment to create final versions of their work product. But developing these skills requires doing the very work that is being outsourced to technology.\u00a0<\/p>\n

Panelists at the Legalweek event wondered how future senior lawyers were being developed \u2014 and about the best courses of action for law firms to consider to safeguard the future of the profession.\u00a0<\/p>\n

On Thursday at ILTACON, just five months later, a new Steno panel provided some detailed answers. The industry has had five months to address these types of concerns, and companies are now showing their work.\u00a0<\/p>\n

Titled \u201cAI as a Teammate: Use Cases That Drive Talent Development,\u201d the discussion focused on ways to utilize AI tools without stymying the development of junior lawyers.\u00a0<\/p>\n

For those who may be worried about the state of lawyer training, the panelists shared some reasons for optimism.\u00a0<\/p>\n

One panelist wondered, for example, whether the training value of traditional lawyer tasks may be overstated.\u00a0<\/p>\n

\u201cIt\u2019s not that there\u2019s no value in going through the paper boxes or the diligence room,\u201d said Nick Hafen, the head of legal technology education at BYU Law School. \u201cBut once you know what a clause looks like, doing it again and again doesn\u2019t add a lot of value to your training.\u201d\u00a0<\/p>\n

\u201cAnd I think AI can concentrate and enhance that training.\u201d\u00a0<\/p>\n

Here, we share some key takeaways from Steno\u2019s portion of the panel. <\/p>\n

Introducing the \u2018Doppelganger\u2019<\/strong><\/p>\n

Since the launch of generative AI, applications that create first drafts of a lawyer\u2019s work product have been most frequently compared to junior associates when describing their function within a firm.\u00a0<\/p>\n

But what if that idea were reframed? What if, instead of creating a generic \u201cassociate,\u201d an AI tool focused on becoming a \u201cdoppelganger\u201d of a specific lawyer you\u2019d want to emulate?\u00a0<\/p>\n

Jeff Cox, an attorney who\u2019s the director of brand at Steno, laid out the case in detail at the ILTACON panel on Thursday. He described an approach that wasn\u2019t 100-feet wide and 1-foot deep, but instead was 1-foot wide and 100-feet deep.\u00a0<\/p>\n

\u201cA generic AI writing system produces output calibrated against a broad model of what \u2018legal writing\u2019 looks like,\u201d he said. \u201cThat can be useful, but it\u2019s not what we\u2019re talking about here.\u201d\u00a0<\/p>\n

Instead, \u201ca \u2018doppelganger\u2019 produces output calibrated against what a specific lawyer\u2019s legal writing actually looks like.\u201d<\/p>\n

In other words, instead of trying to be all things to all people, an AI doppelganger is trained to mimic an elite lawyer \u2014 how they open an argument, how they deal with difficult facts, how they approach a concession so that it doesn\u2019t look like a retreat.\u00a0<\/p>\n

\u201cThat logic is real. It is learnable,\u201d Cox said. \u201cIt is even teachable.\u201d\u00a0<\/p>\n

A Mentorship Boost<\/strong><\/p>\n

When it comes to training junior lawyers, the current mentorship model isn\u2019t exactly perfect.\u00a0<\/p>\n

It requires the right matter, the right assignment, and the right bandwidth. Much of it boils down to a question of access, a commodity that isn\u2019t evenly distributed in law firms, Cox noted.<\/p>\n

What if, instead, the associate had a digital version of a senior partner to interact with? Cox laid out a rundown of the process:<\/p>\n

\u2022 The associate produces an outline or summary of research and sends it to the partner\u2019s digital doppelganger system
\u2022 The system applies feedback, based on its training on the partner\u2019s work
\u2022 The associate studies the response, identifying where the associate\u2019s own instincts diverged from the doppelganger\u2019s
\u2022 The associate finalizes the marked-up draft and takes it to the partner
\u2022 The associate enters this conversation with a hypothesis of what the partner\u2019s feedback will be
\u2022 When the partner and associate meet, the conversation is far more substantive than it would be without this process in advance\u00a0<\/p>\n

The effects of such a system could be profound, according to Cox.\u00a0<\/p>\n

The associate enters the conversation looking to confirm a theory, rather than just submitting the associate\u2019s version of the work product and awaiting surprise feedback. The process encourages, democratizes, and accelerates the mentorship process.<\/p>\n

\u201cIt compresses the feedback loop that mentorship at its best is always trying to create,\u201d Cox said.\u00a0<\/p>\n

Additionally, the digital doppelganger is always available \u2014 and available to all of the lawyers at the firm, not just the associates who work with that partner.<\/p>\n

\u201cIt becomes infrastructure,\u201d Cox said, \u201cavailable without competing for partner time and calendar time.\u201d\u00a0<\/p>\n

Creating the Doppelganger<\/strong><\/p>\n

At many law firms, the traditional process of finalizing a legal document is rife with inefficiencies.\u00a0<\/p>\n

An associate\u2019s draft gets redlined by a partner, often as a hard copy that\u2019s marked up by hand. Several rounds of review or multiple reviewers can create competing drafts. Confusion, lengthy processes, and interpersonal frustrations ensue, before the final work product is ultimately sent out.\u00a0<\/p>\n

It\u2019s a process that can seem comically archaic and inefficient. But it\u2019s also the perfect training ground for a large language model.\u00a0<\/p>\n

A marked up draft reveals a lawyer\u2019s core reasoning, which can be supplemented with analytical memos and other correspondence from that lawyer.\u00a0<\/p>\n

To create the doppelganger, Cox said, you can draw on a variety of materials to show how a lawyer moves a complex matter toward its conclusion.\u00a0<\/p>\n

\u201cWhy did this argument go first?\u201d Cox said. \u201cWhy was this fact treated as background rather than foreground?\u201d<\/p>\n

\u201cWriting out the reasoning behind instructional choices attached to the work that illustrate it is what transforms the voice replica into a genuine thinking partner.\u201d\u00a0\u00a0<\/p>\n

The Cutting Edge<\/strong><\/p>\n

Many like to joke that, when it comes to innovation, law firms are in a \u201crace to be second.\u201d No one wants to be an outlier in a famously risk-averse industry, but everyone does need to keep up.\u00a0<\/p>\n

For law firms, there\u2019s game-changing potential for having the judgment of an excellent lawyer encoded as baseline infrastructure, Cox said. And the time to act is now.\u00a0\u00a0<\/p>\n

\u201cThe lawyers and law firms who build this now will be practicing at a materially different level than those who wait,\u201d Cox said. \u201cI think the question is not whether to do it. The question is how much longer you can afford to wait.\u201d<\/p>\n

Interested in discussing your litigation support options with Jeff and his team? Feel free to <\/em>schedule a Steno demo here<\/em><\/a>.\u00a0<\/em><\/p>\n

The post From \u2018Associate\u2019 To \u2018Doppelganger\u2019: The Evolution Of The Legal AI Persona<\/a> appeared first on Above the Law<\/a>.<\/p>\n

\"\"
\u201cAI as a Teammate\u201d panelists, from left, Jeff Cox of Steno, Damien Riehl of Clio, Daniel Himmel of Legora, and Nick Hafen of BYU Law School.
<\/figcaption><\/figure>\n

If there\u2019s one overarching takeaway from ILTACON 2026, it\u2019s that legal tech has hit the big time<\/a>.<\/p>\n

As Above the Law\u2019s Stephen Embry noted, this year\u2019s conference was CES-like<\/a>, taking over the 700,000-square-foot Gaylord Opryland convention space in Nashville and drawing over 5,700 attendees.\u00a0<\/p>\n

Along with the sheer scale of ILTACON, the event also revealed the industry\u2019s rapid pace of change \u2014 even just over the past few months.\u00a0<\/p>\n

Take a pair of panels featuring Steno, the legal tech company focused on court reporting and litigation support.<\/p>\n

Back in March, Steno was featured in a panel at Legalweek, and one participant \u2014 a technology professional at a major law firm \u2014 was warning attendees of a big problem<\/a> on the horizon.\u00a0<\/p>\n

The problem, he said, was training junior lawyers, and he was worried the industry could be like \u201ca snake eating its own tail.\u201d\u00a0<\/p>\n

That\u2019s because of an apparent paradox: Today\u2019s senior lawyers can vet AI outputs and apply advanced judgment to create final versions of their work product. But developing these skills requires doing the very work that is being outsourced to technology.\u00a0<\/p>\n

Panelists at the Legalweek event wondered how future senior lawyers were being developed \u2014 and about the best courses of action for law firms to consider to safeguard the future of the profession.\u00a0<\/p>\n

On Thursday at ILTACON, just five months later, a new Steno panel provided some detailed answers. The industry has had five months to address these types of concerns, and companies are now showing their work.\u00a0<\/p>\n

Titled \u201cAI as a Teammate: Use Cases That Drive Talent Development,\u201d the discussion focused on ways to utilize AI tools without stymying the development of junior lawyers.\u00a0<\/p>\n

For those who may be worried about the state of lawyer training, the panelists shared some reasons for optimism.\u00a0<\/p>\n

One panelist wondered, for example, whether the training value of traditional lawyer tasks may be overstated.\u00a0<\/p>\n

\u201cIt\u2019s not that there\u2019s no value in going through the paper boxes or the diligence room,\u201d said Nick Hafen, the head of legal technology education at BYU Law School. \u201cBut once you know what a clause looks like, doing it again and again doesn\u2019t add a lot of value to your training.\u201d\u00a0<\/p>\n

\u201cAnd I think AI can concentrate and enhance that training.\u201d\u00a0<\/p>\n

Here, we share some key takeaways from Steno\u2019s portion of the panel. <\/p>\n

Introducing the \u2018Doppelganger\u2019<\/strong><\/p>\n

Since the launch of generative AI, applications that create first drafts of a lawyer\u2019s work product have been most frequently compared to junior associates when describing their function within a firm.\u00a0<\/p>\n

But what if that idea were reframed? What if, instead of creating a generic \u201cassociate,\u201d an AI tool focused on becoming a \u201cdoppelganger\u201d of a specific lawyer you\u2019d want to emulate?\u00a0<\/p>\n

Jeff Cox, an attorney who\u2019s the director of brand at Steno, laid out the case in detail at the ILTACON panel on Thursday. He described an approach that wasn\u2019t 100-feet wide and 1-foot deep, but instead was 1-foot wide and 100-feet deep.\u00a0<\/p>\n

\u201cA generic AI writing system produces output calibrated against a broad model of what \u2018legal writing\u2019 looks like,\u201d he said. \u201cThat can be useful, but it\u2019s not what we\u2019re talking about here.\u201d\u00a0<\/p>\n

Instead, \u201ca \u2018doppelganger\u2019 produces output calibrated against what a specific lawyer\u2019s legal writing actually looks like.\u201d<\/p>\n

In other words, instead of trying to be all things to all people, an AI doppelganger is trained to mimic an elite lawyer \u2014 how they open an argument, how they deal with difficult facts, how they approach a concession so that it doesn\u2019t look like a retreat.\u00a0<\/p>\n

\u201cThat logic is real. It is learnable,\u201d Cox said. \u201cIt is even teachable.\u201d\u00a0<\/p>\n

A Mentorship Boost<\/strong><\/p>\n

When it comes to training junior lawyers, the current mentorship model isn\u2019t exactly perfect.\u00a0<\/p>\n

It requires the right matter, the right assignment, and the right bandwidth. Much of it boils down to a question of access, a commodity that isn\u2019t evenly distributed in law firms, Cox noted.<\/p>\n

What if, instead, the associate had a digital version of a senior partner to interact with? Cox laid out a rundown of the process:<\/p>\n

\u2022 The associate produces an outline or summary of research and sends it to the partner\u2019s digital doppelganger system
\u2022 The system applies feedback, based on its training on the partner\u2019s work
\u2022 The associate studies the response, identifying where the associate\u2019s own instincts diverged from the doppelganger\u2019s
\u2022 The associate finalizes the marked-up draft and takes it to the partner
\u2022 The associate enters this conversation with a hypothesis of what the partner\u2019s feedback will be
\u2022 When the partner and associate meet, the conversation is far more substantive than it would be without this process in advance\u00a0<\/p>\n

The effects of such a system could be profound, according to Cox.\u00a0<\/p>\n

The associate enters the conversation looking to confirm a theory, rather than just submitting the associate\u2019s version of the work product and awaiting surprise feedback. The process encourages, democratizes, and accelerates the mentorship process.<\/p>\n

\u201cIt compresses the feedback loop that mentorship at its best is always trying to create,\u201d Cox said.\u00a0<\/p>\n

Additionally, the digital doppelganger is always available \u2014 and available to all of the lawyers at the firm, not just the associates who work with that partner.<\/p>\n

\u201cIt becomes infrastructure,\u201d Cox said, \u201cavailable without competing for partner time and calendar time.\u201d\u00a0<\/p>\n

Creating the Doppelganger<\/strong><\/p>\n

At many law firms, the traditional process of finalizing a legal document is rife with inefficiencies.\u00a0<\/p>\n

An associate\u2019s draft gets redlined by a partner, often as a hard copy that\u2019s marked up by hand. Several rounds of review or multiple reviewers can create competing drafts. Confusion, lengthy processes, and interpersonal frustrations ensue, before the final work product is ultimately sent out.\u00a0<\/p>\n

It\u2019s a process that can seem comically archaic and inefficient. But it\u2019s also the perfect training ground for a large language model.\u00a0<\/p>\n

A marked up draft reveals a lawyer\u2019s core reasoning, which can be supplemented with analytical memos and other correspondence from that lawyer.\u00a0<\/p>\n

To create the doppelganger, Cox said, you can draw on a variety of materials to show how a lawyer moves a complex matter toward its conclusion.\u00a0<\/p>\n

\u201cWhy did this argument go first?\u201d Cox said. \u201cWhy was this fact treated as background rather than foreground?\u201d<\/p>\n

\u201cWriting out the reasoning behind instructional choices attached to the work that illustrate it is what transforms the voice replica into a genuine thinking partner.\u201d\u00a0\u00a0<\/p>\n

The Cutting Edge<\/strong><\/p>\n

Many like to joke that, when it comes to innovation, law firms are in a \u201crace to be second.\u201d No one wants to be an outlier in a famously risk-averse industry, but everyone does need to keep up.\u00a0<\/p>\n

For law firms, there\u2019s game-changing potential for having the judgment of an excellent lawyer encoded as baseline infrastructure, Cox said. And the time to act is now.\u00a0\u00a0<\/p>\n

\u201cThe lawyers and law firms who build this now will be practicing at a materially different level than those who wait,\u201d Cox said. \u201cI think the question is not whether to do it. The question is how much longer you can afford to wait.\u201d<\/p>\n

Interested in discussing your litigation support options with Jeff and his team? Feel free to <\/em>schedule a Steno demo here<\/em><\/a>.\u00a0<\/em><\/p>\n

The post From \u2018Associate\u2019 To \u2018Doppelganger\u2019: The Evolution Of The Legal AI Persona<\/a> appeared first on Above the Law<\/a>.<\/p>\n","protected":false},"excerpt":{"rendered":"

\u201cAI as a Teammate\u201d panelists, from left, Jeff Cox of Steno, Damien Riehl of Clio, Daniel Himmel of Legora, and Nick Hafen of BYU Law School. If there\u2019s one overarching takeaway from ILTACON 2026, it\u2019s that legal tech has hit the big time. As Above the Law\u2019s Stephen Embry noted, this year\u2019s conference was CES-like, […]<\/p>\n","protected":false},"author":3,"featured_media":162208,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"_et_pb_use_builder":"","_et_pb_old_content":"","_et_gb_content_width":"","_jetpack_newsletter_access":"","_jetpack_dont_email_post_to_subs":false,"_jetpack_newsletter_tier_id":0,"_jetpack_memberships_contains_paywalled_content":false,"_jetpack_memberships_contains_paid_content":false,"footnotes":""},"categories":[16],"tags":[],"class_list":["post-162224","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-above_the_law"],"jetpack_sharing_enabled":true,"jetpack_featured_media_url":"https:\/\/i0.wp.com\/xira.com\/p\/wp-content\/uploads\/2026\/09\/IMG_9952-1024x570-pi6wjO.jpg?fit=1024%2C570&ssl=1","_links":{"self":[{"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/posts\/162224","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=162224"}],"version-history":[{"count":0,"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/posts\/162224\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/media\/162208"}],"wp:attachment":[{"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/media?parent=162224"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/categories?post=162224"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/tags?post=162224"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}