# The Tower of Babel in Your Own Building Why Lawyers Speak Latin, IT Speaks Geek, and Neither Side Translates Before the Sanctions Hit THE TECHNOLOGY BLIND SPOT -A Blog Series on Legal Technology, Cybersecurity, and Professional Responsibility Magistrate Judge Barbara Major used nine words to end six legal careers: “assisted Qualcomm in committing this incredible discovery violation.” Six attorneys. $8.5 million in sanctions. A referral to the California State Bar. And 46,000 documents sitting on employee hard drives the entire time. Unproduced. Unsearched. Invisible to the litigation team. Nobody on the legal side understood where the data lived in Qualcomm’s IT infrastructure. Nobody on the IT side understood why it mattered. This is not a story about one company’s discovery failure. This is a story about the Technology Blind Spot: the systematic gap between what technology professionals do and what legal professionals believe they do. Two professions built on precision. Operating in the same building. Working against the same deadline. Speaking two languages that share no common vocabulary. Legal professionals inherit 900 years of Latin maxims, common law precedent, and Rules of Professional Conduct. IT professionals operate in an equally precise dialect of Application Programming Interfaces (APIs), cloud architectures, hash values, and system administration protocols. Both professions demand that words mean exactly what they say. When these two groups must collaborate, the precision that defines each profession becomes the barrier that prevents communication between them. The cost of this barrier is measurable. IDC research published in 2025 found that businesses lose an estimated 11 percent of annual revenue to legal friction, amounting to $141 million per year for a mid-to-large organization. Clio’s 2024 Legal Trends Report documented that the average lawyer bills 2.5 hours of an eight-hour day. Some of those lost hours are irreducible overhead. The portion attributable to two departments failing to communicate about the same deadline is not overhead. It is a translation failure with a price tag. ## The Copy That Was Not a Copy Before examining how the Qualcomm disaster unfolded, consider the quieter version of the same failure that repeats in law firms every week. Litigation is anticipated. A legal hold is issued. The litigation partner sends an urgent email to the IT director: “We need a forensically sound image of the custodian’s device before close of business. Preserve the metadata. Chain of custody is paramount.” The IT director reads the message, nods, and does what IT professionals do when asked to “copy” something. He drags the folders to a network share. Every byte arrives intact. A standard copy operation, whether executed through Windows Explorer or a command-line transfer, changes the metadata on every file it touches. The “last accessed” timestamp overwrites. The “date created” field reflects the copy date, not the original creation date. In some configurations, the “last modified” date shifts as well. File content remains identical. The file’s digital provenance, the information establishing when the document was created, who accessed it, and when, is altered or destroyed. Metadata is Electronically Stored Information (ESI) governed by FRCP Rules 26 and 34. Courts require proof that electronic evidence matches the ESI originally collected. Without intact metadata, authentication demands sworn testimony that would otherwise be unnecessary, and opposing counsel gains a credible basis to challenge the entire production. ## Exhibit: What “Forensically Sound” Actually Means A forensic image creates a bit-for-bit duplicate of the entire storage device using write-blocking hardware or software that prevents any modification to the source. The image includes a cryptographic hash value. The Electronic Discovery Reference Model (EDRM) defines the standard: “A forensic image is a sector-by-sector copy of the media that replicates the source completely, including deleted data, file slack, and unallocated space, and is verified by comparing hash values of the source and the copy.” That hash, typically SHA-256, functions as a digital fingerprint. If even a single bit changes after collection, the hash value changes, providing mathematically verifiable proof that the evidence is unaltered. This is what the litigation partner meant by “forensically sound.” This is not what the IT director understood. The partner said “forensically sound image.” The IT director heard “make a copy.” Neither was wrong in their own language. Both were catastrophically wrong in each other’s. The consequences include spoliation findings, adverse inference instructions, monetary sanctions, and in extreme cases, default judgments. Once metadata is spoiled, it cannot be reconstructed. The error is permanent, the prejudice is presumed, and the sanctions exposure falls on the attorney who signed the Rule 26(g) certification. ## When the Tower Fell: Qualcomm v. Broadcom If the forensic copy scenario illustrates the daily cost of mistranslation,Qualcomm Inc. v. Broadcom Corp.(S.D. Cal. 2008) illustrates the catastrophic version. During a patent dispute, Broadcom sought documents related to Qualcomm’s participation in a video coding standards-setting body called the Joint Video Team (JVT). Qualcomm produced 1.2 million pages. Trial counsel argued Qualcomm had not participated in the JVT, and therefore had not waived its patent claims. Under cross-examination, a Qualcomm witness revealed she had received emails from the JVT mailing list. Counsel had told the judge these emails did not exist. After the jury found for Broadcom, Qualcomm’s own general counsel conducted basic keyword searches on computers the litigation team had never examined. Those searches revealed more than 46,000 emails that directly undermined the arguments counsel had made at trial. The documents had been there all along, stored on custodian hard drives. Nobody searched them because the attorneys did not understand where the data lived, and the IT team was never asked the right questions. Judge Major’s language left no room for ambiguity. The court’s order stated: “One or more of the retained lawyers chose not to look in the correct locations for the correct documents, to accept the unsubstantiated assurances of an important client that its search was sufficient, to ignore the warning signs that the document search and production were inadequate, not to press Qualcomm employees for the truth, and/or to encourage employees to provide the information (or lack of information) that Qualcomm needed to assert its non-participation argument and to succeed in this lawsuit.” The court imposed $8,568,633.24 in sanctions and referred six attorneys to the State Bar of California. The court held that the attorneys “may have violated their ethical duties” and found it “unbelievable” that they did not “know or suspect” the inadequacy of the search. On remand in 2010, after Qualcomm waived privilege and allowed examination of internal communications, the magistrate judge dissolved the sanctions against the individual attorneys. The finding was not that the attorneys performed adequately, but that privileged communications shifted responsibility to the client. By then, the damage was irreversible. Several attorneys abandoned large-firm practice. Careers built over decades ended not because of malice, but because of a communication failure between professionals who never learned to speak each other’s language. I have watched this pattern before. During my years at Dell EMC, I saw the same breakdown between engineering teams and enterprise sales organizations. Engineers spoke in RAID configurations and storage architectures. Sales teams spoke in total cost of ownership. When neither side translated, customers received systems that met every technical specification and solved none of their business problems. The pattern is identical. In enterprise technology, that cost us quarterly revenue. In legal practice, it costs careers and bar licenses. ## The Structural Mismatch Beyond vocabulary, the gap reflects a difference in professional DNA. Lawyers train to minimize risk, preserve precedent, and operate within adversarial frameworks. IT professionals train to iterate, deploy, test, fail, and optimize. Legal instinct: slow down and get it right the first time. IT instinct: move fast and improve continuously. A Bloomberg Law survey of 936 practicing attorneys found that legal technology and data literacy ranked well below traditional lawyering skills in perceived importance. Meanwhile, Clio’s 2024 Legal Trends Report documented that AI adoption among lawyers surged from 19 percent in 2023 to 79 percent in 2024. Attorneys are adopting tools they were never trained to govern, implemented by IT teams they were never trained to communicate with. This disconnect, as I documented in “AI Won’t Wait for Lawyers to Get Comfortable”, follows a pattern visible across every industry I have worked in: adoption outpaces governance, governance outpaces communication, and the gap between the first and last produces the sanctions docket. ## Jargon as Identity, Precision as Barrier Both professions use specialized language for legitimate reasons. Legal Latin persists because terms like “res judicata,” “habeas corpus,” and “mens rea” carry centuries of judicial interpretation that no English paraphrase replicates with equal certainty. When a lawyer says “res judicata,” every lawyer in the room understands not just the concept but the entire body of case law defining its boundaries. The term is compression, not pretension. IT jargon operates on the same principle. When an engineer says “SHA-256 hash,” that term compresses the full specification of a one-way cryptographic function producing a unique 256-bit fingerprint for any digital file. Precise because the technology demands it. Both vocabularies simultaneously unify members of a professional group and exclude outsiders. The exclusion is not always intentional. It is always consequential. When a litigation partner says “preserve the metadata,” the partner means “maintain the evidentiary integrity of every attribute attached to every file so that chain of custody is defensible under the Federal Rules of Evidence.” When the IT director hears “preserve the metadata,” the director thinks “don’t rename the files.” Both are technically correct definitions. The distance between them is measured in sanctions. ## The Ethical Floor: What the Rules Actually Require The ABA has not been silent on this problem, though the profession has been slow to listen. And the ABA’s own framing reveals the depth of the failure. When the ABA amended Comment 8 in 2012, its Committee Report to the House of Delegates stated the change “does not impose any new obligations” on lawyers. The duty to understand the technology integral to client representation was not new in 2012. It had always existed under Rule 1.1. The ABA was issuing a reminder. Yet 40 states needed over a decade to adopt even that reminder, and as of 2023 only eight percent of law students reported taking a legal technology class. The profession’s own regulatory body conceded the obligation predated the amendment. The profession’s own training pipeline still does not meet it. Comment 8 requires lawyers to “keep abreast of changes in the law and its practice, including the benefits and risks associated with relevant technology.” This does not require becoming technologists. It requires understanding technology well enough to supervise it, which means understanding the people who operate it. Rule 1.6(c) adds a parallel obligation: “reasonable efforts” to prevent unauthorized disclosure of client information, efforts that require understanding how client data is stored, transmitted, and preserved. As I analyzed in “Your AI Tool Doesn’t Keep Secrets”, the “reasonable efforts” standard breaks down where legal instructions meet technical execution. Supervisory obligations under Rule 5.1 close the loop. If a supervising attorney delegates eDiscovery collection to an IT team without ensuring the team understands evidentiary standards, the supervising attorney bears responsibility for any resulting deficiency. You cannot supervise what you cannot communicate. Formal Opinion 483 (2018) makes this explicit after a data breach: lawyers must understand enough about their firm’s technical infrastructure to evaluate what happened and what data may have been compromised. The opinion does not, however, provide guidance on how to conduct those conversations. That gap is where the Tower of Babel stands. ## The Skeptic’s Objection Two predictable objections deserve direct engagement. First, that lawyers have always been technology-averse and always will be. The data destroys this argument. AI adoption among lawyers surged from 19 percent to 79 percent in a single year. Nearly 90 percent of respondents to Bloomberg Law’s Legal Ops and Tech Survey said legal technology is important to meeting client demands. The profession is not technology-averse. It is translation-averse. Lawyers adopt tools at remarkable speed when the value proposition is clear. What they resist is the cognitive load of learning a second professional language without guidance or institutional support. Second, that firms should simply hire translators: legal engineers, legal operations professionals, or eDiscovery specialists who straddle both worlds. The ABA endorsed this approach in a 2025 Law Technology Today article, noting that legal engineers “serve as translators between lawyers and developers, ensuring the tools being built actually meet legal teams’ needs.” This is a valid tactical response. It does not address the structural problem. Translation intermediaries help on specific projects. They do not fix the organizational silos that created the need for translation. ## Where This Argument Is Weakest An honest assessment of this piece’s own limitations is necessary before prescribing solutions. Qualcomm was decided in 2008. The eDiscovery landscape has matured considerably since then. Sophisticated litigation practices now routinely engage specialized eDiscovery vendors who manage forensic collection, processing, and review. For AmLaw 200 firms handling complex commercial litigation, the specific gap described in the Qualcomm scenario, attorneys failing to search custodian hard drives, would be unusual today. That concession is real, and it does not save the skeptic’s position. The communication gap has not closed. It has migrated. In 2008, the gap lived in forensic collection. In 2026, it lives in AI tool governance, cloud security configurations, and data residency implications of SaaS platforms processing client information across jurisdictions. Catherine’s firm may no longer bungle a forensic image. Her firm almost certainly cannot articulate where client data resides in its AI-assisted document review platform or what the vendor’s subprocessor chain looks like. The Tower of Babel rebuilt itself on a different floor. ## Bridging the Gap: What Firms Can Do This Week The path forward requires action at three levels: individual competency, organizational structure, and technological infrastructure. Individual Competency.Every litigator who may touch eDiscovery should understand the difference between a standard copy and a forensic image, the role of hash values in chain-of-custody documentation, and the metadata fields courts examine when evaluating ESI authenticity. Every IT professional supporting legal operations should understand what a litigation hold requires, why metadata integrity matters for evidentiary purposes, and what “forensically sound” means under FRCP Rule 37(e). The concrete first step: schedule a 90-minute cross-training session this month where your lead litigator and IT director each present their five most critical terms, with definitions written for the other profession’s vocabulary. Organizational Structure.Legal and IT cannot remain siloed departments that communicate only during crises. Both departments understand “reduce time-to-billing by 30 percent” and “avoid a seven-figure sanctions order.” Start with shared outcomes, not shared vocabulary. Build a glossary of the 20 terms most likely to cause misunderstanding in your firm’s practice areas: “forensically sound,” “metadata,” “litigation hold,” “write-blocker,” “chain of custody,” “hash value.” Define each term twice, once for lawyers and once for IT. Post it where both teams can reach it before the next discovery deadline arrives. Technological Infrastructure.Technology itself is becoming the translation layer. Natural language interfaces, agentic AI systems, and purpose-built legal operations platforms reduce the need for either side to master the other’s jargon by embedding legal logic into workflows. KPMG projects that by 2030, a legal professional’s ability to use technology will become a key differentiator. The firms investing in this translation layer now will define the competitive standard. The firms that wait will explain to their partners why the sanctions motion cited their collection protocol. ## The Language of Consequences Judge Major ordered six attorneys to forward her opinion to the State Bar. She expressed hope that the case would “establish a turning point” in the profession’s approach to discovery obligations. That was 2008. Seventeen years later, the turning point has not arrived. The vocabulary has changed. The structural failure has not. The Qualcomm attorneys did not fail because they were lazy or dishonest. They failed because they never learned where the data lived. The IT director who made a standard copy instead of a forensic image did not fail because he was careless. Nobody explained the legal difference between a copy and evidence. Both groups did exactly what their training prepared them to do. Their training never prepared them for each other. U.S. legal services contribute approximately $396 billion annually to the national economy. The portion of that figure lost to two departments in the same building speaking two different languages about the same deadline is not a cost of doing business. It is a choice. And choices, unlike metadata, can still be changed. This blog provides general information for educational purposes only and does not constitute legal advice. Consult qualified counsel for advice on specific situations. ## About the Author JD Morris is Co-Founder and COO of LexAxiom. With over 20 years of enterprise technology experience and credentials including an MLS from Texas A&M, MEng from George Washington University, and dual MBAs from Columbia Business School and Berkeley Haas, JD focuses on the intersection of legal technology, cybersecurity, and professional responsibility. Connect:LinkedIn|X|Bluesky ## References Qualcomm Inc. v. Broadcom Corp., No. 05cv1958-B (BLM), 2008 WL 66932 (S.D. Cal. Jan. 7, 2008) Qualcomm Inc. v. Broadcom Corp., No. 05cv1958-B (BLM) (S.D. Cal. Apr. 2, 2010) (remand decision dissolving attorney sanctions) ABA Model Rule 1.1, Comment 8 (Technological Competence) ABA Model Rule 1.6(c) (Confidentiality / Reasonable Efforts) ABA Model Rule 5.1 (Supervisory Duties) ABA Formal Opinion 477R (2017) (Securing Electronic Communications) ABA Formal Opinion 483 (2018) (Post-Breach Obligations) ABA Commission on Ethics 20/20, Report to the House of Delegates (August 2012) (stating Comment 8 amendment “does not impose any new obligations”) Robert Ambrogi, “Tech Competence,” LawSites (tracking 40+ state adoptions of Comment 8, updated 2025) Fed. R. Civ. P. 26, 34, 37(e) (ESI Discovery Obligations) Bloomberg Law, Spring 2023 Law School Preparedness Survey (936 practicing attorneys) Bloomberg Law, 2022 Legal Ops + Tech Survey Clio, Legal Trends Report 2024 (AI adoption data: 19% to 79%) IDC White Paper, sponsored by LawVu, “Legal Friction: The Real Cost to Your Business,” #AP15041X (April 2025) U.S. Bureau of Economic Analysis, Gross Domestic Product: Legal Services (NAICS 5411), $387.7 billion (2024) EDRM (Electronic Discovery Reference Model), Collection and Forensic Imaging Standards KPMG, “10 Predictions: The Legal Department of the Future” (2025) ABA Law Technology Today, “How Legal Engineers Are Reshaping Modern Law Firms” (2025) Morris Legal Technology Blog, “AI Won’t Wait for Lawyers to Get Comfortable” (2025) Morris Legal Technology Blog, “Your AI Tool Doesn’t Keep Secrets” (2025)

Originally published on LinkedIn Newsletter: The Technology Blind Spot