AI in Legal Practice

3 Workflows I've Automated for in-house teams. ① Ask Legal ② Procurement ③ Contract Review (not just the review!) 1. Ask Legal [or any department for that matter 🤷🏼♀️] You've heard me talk about legal teams and knowledge management. Long story short, your legal team is answering the same 20 questions over and over 😵💫 A simple way to save a CHUNK of time answering questions from the business (enabling them to go faster) ALL while having complete control & keeping a human in the loop? ↪️ Set up an 'Ask Legal' bot in your comms platform. ↪️ Sync it with your knowledge base (e.g GDrive/Notion/Sharepoint). ↪️ Set up your custom instructions (Want it to tag Bob on privacy questions only, specifically on a Tuesday? No problem).  ↪️ Don't want the answer to go straight out to the business without reviewing it first? Cool, turn on co-pilot mode. The result? 60-80% fewer repetitive queries. Your team focuses on the high value things that need a human lawyer. 2. Procurement Businesses have 100's of tools, but when departments don't speak to each other you end up with duplicate tools & subscriptions 😭 💵 🚽.  What if there was a way for the business to find out in <1 minute if there was a tool available that covered their needs, before needing to spend some hard secured department budget? Moreover, what if I told you, they could kick off the internal procurement process from the comfort of your comms platform? Team member : “Do we already have a tool for X?” in Slack/Teams ✅ Bot checks knowledge base (policies, procurement tool). ✅ If a match is found, it shares the approved tool & owner to contact. ✅ If not, the bot can ask the user for more info and direct them with next steps to kick off the procurement process from inside Slack/Teams. Ensuring your users ACTUALLY follow the process, without adding friction. Did I just see your CFO cry tears of joy? 3. Third Party Vendor Contract Review & Project Management Getting AI to redline a contract (as a first pass) is a huge win, but there's still the other pieces of the process missing, like: 🤷🏼♀️ The business figuring out IF legal review is even needed (according to company policy). 📨 The business actually submitting the contract to legal. 😩 Managing review capacity within the legal team. 🖥️ Getting the legal team to log & update the PM tool. The list never ends. Legal reviews only what actually needs their eyes, turnaround times improve, and the business stops pinging the team for “update pls?” in Slack : ) TLDR; Most legal teams are drowning in admin work that could be automated. I've built all of these using simple processes and tools (that I've found most businesses have). You also know I love a good Figma flow. So I’ve built them for all three of the above (see a sneak peak below). Want the entire thing? Comment "FLOWS" and I'll send them over. Also, tell me what you want to see - more of the above or step-by-step how-to build videos?

Legal AI isn’t coming. It’s already a full market map. Contracts, research, negotiations, patents – that’s the engine of legal work. It’s dense and high-risk. And AI helps navigate it with remarkable efficiency. Here’s a simple map of where AI fits: 1. Contract Management Tracks obligations, deadlines and risky clauses across large contract portfolios. What this changes: missed details cost money. 2. Legal Research & Intelligence Scans case law and regulations in seconds. What this changes: better research, stronger decisions. 3. Document Drafting & Negotiation Generates drafts and compares versions during negotiations. What this changes: less formatting, more strategy. 4. Workflow-Specific Tools Automates tasks in compliance, litigation, due diligence, e-discovery. What this changes: faster processes, fewer errors. 5. IP Management Manages trademarks, copyrights, and patent portfolios. What this changes: IP is often a company’s biggest asset. 6. IP Protection & Monetization Monitors infringements and identifies licensing opportunities. What this changes: protect ideas and turn them into revenue. 7. Patent Drafting & Review Supports precise claim writing and consistency checks. What this changes: small wording gaps can weaken protection. 8. Patent Research Searches global databases for prior art. What this changes: avoids costly rejections. Let machines scan. And let humans decide. That balance is exactly where platforms like e! by Lexemo come in. Not “AI for everything,” but structured legal automation where AI is used deliberately and transparently. If you’re thinking about automating legal processes, take a look: https://lnkd.in/g49NngZn Have you ever had situations where AI helped you sort out a legal mess? Or avoid one altogether? #LegalTech #Automation #AutoMate

If I didn't tell you this was Claude, you'd think a Legal AI company just launched a new product. Look at these features: → Work through comments — reads each one, edits the anchored text as a tracked change, replies with what it did → Draft in your template — writes in your heading and bullet styles, with citations back to source docs → Check for consistency — flags inconsistent defined terms, broken cross-references, numbering errors → Review revisions — reads tracked changes as proposed edits for you to accept or reject And the real kicker: Turn workflows in Word into skills. Your team nails a contract review process? Save it as a reusable skill. Anyone on the team can run it, same quality output. When Claude first launched its legal plugin, I published my Legal AI Value Stack. Back then, the plugin impressed me but still felt like a preview. Today? This looks like a complete workflow product. 🔴 Level 1 (Raw AI Capability) — already commoditized 🟡 Level 2 (AI + Workflow/UI) — this is what $300/seat/month legal AI startups sell. And Claude just shipped it as a feature, not a product. If you're entering legal AI today, my honest advice: start at Level 3 and above. Generic contract review as a workflow play seems over. 🟢 Level 3: Proprietary Data — data foundation models can't access 🏰 Level 4: System of Record — mission-critical infrastructure with high switching costs 🚀 Level 5: Hybrid Model — AI + human legal service combined at scale Or — if you're building at Level 2, go deep into a vertical where general-purpose AI can't follow (maybe?): patent prosecution, immigration, litigation... The moat isn't the workflow anymore. It's the domain depth underneath it. Detailed breakdown of all 5-level analysis in the comments 👇

Three major developments in the last week should have every HR leader, employer, and AI vendor paying attention: 1. The AI Civil Rights Act was reintroduced in the US Congress Led by Senator Ed Markey and Representative Yvette D. Clarke, this legislation places hard guardrails around AI and algorithmic systems used in decisions related to hiring, housing, healthcare and beyond. It demands transparency, bias testing, and accountability. Think of it as GDPR for bias, but with broader implications across HR, tech, and operations. “We will not allow AI to stand for Accelerating Injustice.” – Senator Ed Markey for U.S. Senate 2. California’s new workplace AI discrimination laws are now in effect. The new rule governing companies' use of automated decision-making technology will likely create a situation where companies are liable for hiring practices if a system violates anti-discrimination laws. As other U.S. states also implement laws and regulations containing similar ADMT protections, companies deploying the technology will need to be proactive in their record keeping and vetting of third-parties while auditing their own tools to understand how the software functions. It’s no longer enough to trust your tools and vendors, you must prove they’re fair. 3. Insurers are backing away from covering AI risks AIG, Great American, and WR Berkley are asking regulators to exclude AI-related liabilities from their policies. Why? Because the risks (from chatbots hallucinating to algorithmic bias in hiring) are seen as “too opaque, too unpredictable.” When insurers are pulling cover, it’s a warning sign: you own the risk. 👁 What this means for HR and recruitment business leaders: We’ve officially entered the age of AI Accountability. That means: ✅ You need visibility into how your AI systems work, especially if they’re used for hiring, performance management, or workforce planning. ✅ You must audit your HR tech stack (yes, that includes Workday, ATS platforms, and even AI resume screeners). ✅ You need to document fairness, not just assume it. ✅ You must rethink your contracts with AI vendors. If the tech goes wrong, insurers may not have your back. 🛡 If you haven’t already, it’s time to start building your AI Governance Playbook. 📌 Audit all AI tools in use 📌 Build an internal AI ethics committee 📌 Ensure legal, DEI and HR alignment on tool deployment 📌 Partner only with vendors offering bias mitigation, auditability, and indemnification

Most lawyers are using AI to do faster what they were already doing. That is not the opportunity. AI agents can complete entire workflows autonomously. Here are three examples from legal work: 1) Lease abstraction at scale. An agent reviews 400 commercial leases, extracts rent escalation clauses, flags deviations from the negotiated form, and outputs a variance report. Before a human opens a single document. What took a team two weeks now takes two hours. 2) Regulatory change management. Instead of a paralegal manually checking state-by-state privacy law updates, an agent monitors legislative feeds, maps changes to existing data processing agreements, and drafts a memo flagging the ones that require action. 3) Deal room diligence. In an M&A transaction, an agent ingests the virtual data room, surfaces missing representations, identifies indemnification gaps, and cross-references disclosed litigation against public court records. Autonomously. What all three share: a human sets the objective, verifies the data, and reviews the output. The agent handles every step in between. The professional responsibility question this raises is not whether to use these tools. It is how to structure meaningful supervision when you are reviewing an agent's work product rather than directing it step by step. That is the question lawyers need to be asking right now. I'm Colin, General Counsel at Malbek and author of The Legal Tech Ecosystem. #legaltech #contracts #law #business #learning

How we helped the legal team at CluePoints build multiple AI workflows in two days.   The team had: - Experimented with ChatGPT Enterprise with no real success - Built a Custom GPT that produced unreliable output - Tried Microsoft Copilot with mixed results They didn't need another demo. They needed structured expertise to go from experimentation to reliable workflows. We designed a four-part programme: - Two live virtual sessions covering AI fundamentals and prompting for legal work - Two half-day in-person AI build workshops in London The virtual sessions laid the groundwork — how LLMs actually work under the hood, prompting patterns and practices tailored to in-house legal practice, and a diagnosis of why AI experiments fall short. Then we built. Day One. We took their actual procurement playbook - real clauses, real fallback positions - and turned it into working review workflows in both Claude CoWork and Microsoft Copilot. Two teams, two tools, both with functioning prototypes by end of day. Day Two. We built a Claude CoWork skill for document review and redlining that cut review time by 75%. We created a chatbot from their internal knowledge base using Copilot Studio. We ran an AI use case roadmap exercise with the team, prioritising what would make a difference to their day-to-day, with ambitious self-imposed efficiency targets to hit by year-end. Alice Sahba, their VP of Legal, nailed it: "It's that prompting forces you to map your legal process precisely enough that a machine can follow it consistently." The team left with skills they own, a roadmap they built, and the ability to keep iterating without us. That's how we work with in-house legal teams. Build with them, not for them. We run this as a structured programme for in-house legal and compliance teams - live virtual AI foundations followed by hands-on build workshops tailored to your contracts, playbooks, and tools. What's the one legal workflow your team would automate first? Drop it in the comments - I'll share how we'd approach it.

🚨 Candidate sues over AI hiring bias — and the case is moving forward. Employers, take note. A federal judge recently allowed a nationwide collective action lawsuit to proceed against Workday, one of the most widely used AI-driven hiring platforms. The case was brought by Derek Mobley, a Black job seeker over 40 with anxiety and depression, who alleges Workday’s algorithms disproportionately screened out candidates based on race, age, and disability status — violating federal anti-discrimination laws. Why this case matters: ✅ It’s not just about one company. This ruling opens the door for thousands of job seekers to join the lawsuit — and puts every employer using AI in recruitment on alert. ✅ Employers may be liable for vendor bias. Even if you’re not building the AI, you’re responsible for how it impacts your hiring outcomes. ✅ Disparate impact claims are real. Seemingly neutral filters like resume gaps or job hopping may unfairly disadvantage protected groups — and that’s actionable. ✅ AI doesn’t erase legal responsibility. If it’s screening candidates, it must be tested for bias, monitored for fairness, and used with human oversight. If you’re in HR, legal, or talent acquisition, here’s what you should be doing now: • Audit your hiring tools — know what’s automated and how decisions are made • Require transparency and bias testing from AI vendors • Retain human oversight in every hiring decision • Track hiring outcomes by race, age, gender, disability — and follow up on disparities • Build an internal AI governance plan that includes HR, Legal, and IT This lawsuit is a warning shot. The future of hiring may be automated — but compliance, fairness, and ethics must remain human-led. #HR #AI #Hiring #EmploymentLaw #Workday #PeopleAnalytics #FairHiring #AIBias #DEI #FutureOfWork

Your job applicant sorting software may be automated decision making that may be prohibited in the EU and may require a bunch of things like a #DPIA and an opt out in the US - even if a hiring counsellor is making the final hiring decision! A new decision from the - per new decision from the Supreme Administrative Court of Austria follows the footsteps of the Schufa decision. In this case the controller, the Public Employment Service in Austria, used an algorithm to calculate the degree of probability for jobseekers to be employed for a certain number of days, based on: (1) age group, (2) gender, (3) country group, (4) education, (5) health impairment, (6) care responsibilities, (7) occupational group, (8) career history and (9) the regional labor market situation and the duration of cases at the controller. Based on this, the algorithm divided jobseekers into the following three groups: (1) Service jobseekers with high labor market opportunities, (2) Care jobseekers with low labor market opportunities, (3) Consultancy jobseekers with medium labor market opportunities. The result was used as a starting point for counsellors to work with jobseekers to assess their potential and any obstacles in the labour market integration. The algorithm itself was not used for job placement, but only for targeted support and assistance, Per the court: 🔹 The algorithm decided on the allocation of jobseeker’s group and thus has a legal effect on the jobseekers concerned or similarly significantly affects them. 🔹 The fact that the final decision on the jobseeker’s group assignment lies with the counsellor, does not prevent the algorithm from being classified as an automated decision under Article 22(1) GDPR. 🔹 The instructions and trainings that were provided to ensure counsellors would not accept the algorithm’s results unquestioningly could not exclude the possibility that the algorithm is ultimately decisive for the allocation. Really important in the US as well since automated decisions that affect the prospect of employment are considered "legal or similarly significant effects" under (most if not all) US State Privacy Laws. #dataprivacy #dataprotection #privacyFOMO #AIprivacy photo by vectorjuice for Freepik https://lnkd.in/eerj7SgW

Two years ago the AI employment lawsuit was about who gets hired. This week it is about who gets fired. 26 Meta employees sued this week, alleging the company used a set of AI tools to score and rank staff for layoffs, and that the system swept up people on medical and parental leave at higher rates. The logic writes itself. If your ranking rewards output, AI-token usage, and "AI-native" activity, someone on leave cannot generate any of it. Absent data becomes a low score. A low score becomes a layoff. Meta's response: workforce decisions were made by people, not AI. That sentence should sound familiar. It is close to what Workday argued in Mobley v. Workday, the case over AI screening out job applicants. Workday called its software a neutral tool that only ran the employer's criteria. Neither case is decided, and Workday has not been found liable. But a judge already refused to dismiss the claims, finding the software may have helped shape the hiring decision rather than only running instructions. AI now sits on both ends of the employment lifecycle, the offer and the exit. Courts have not ruled that "a human clicked the button" fails as a defense. They have ruled it is not enough to end the case early. When the ranking that human relied on was built by a model, the button is no longer where the decision was made. Three things every leader using these tools should check before the next reduction: 𝟭. 𝗔𝘀𝗸 𝘄𝗵𝗮𝘁 𝘆𝗼𝘂𝗿 𝗯𝘂𝘀𝗶𝗻𝗲𝘀𝘀 𝗺𝗼𝗱𝗲𝗹 𝗿𝗲𝘄𝗮𝗿𝗱𝘀. If the inputs are things a person on protected leave physically cannot produce, you have a disparate-impact problem already forming. 𝟮. 𝗞𝗲𝗲𝗽 𝘁𝗵𝗲 𝗵𝘂𝗺𝗮𝗻 𝗱𝗲𝗰𝗶𝘀𝗶𝗼𝗻 𝗿𝗲𝗮𝗹. Reviewing an AI-generated stack rank and approving it is not judgment. It is a rubber stamp with your name on it. 𝟯. 𝗔𝘂𝗱𝗶𝘁 𝘁𝗵𝗲 𝗼𝘂𝘁𝗽𝘂𝘁, 𝗻𝗼𝘁 𝘁𝗵𝗲 𝗶𝗻𝘁𝗲𝗻𝘁. Courts look at who got cut, not at whether you meant well. AI did not invent discrimination in hiring or firing. It scaled a decision we were already making and removed the friction that used to make us pause. 💡 𝗧𝗵𝗲 𝗻𝗲𝘅𝘁 𝘄𝗮𝘃𝗲 𝗼𝗳 𝘁𝗵𝗲𝘀𝗲 𝘀𝘂𝗶𝘁𝘀 𝘄𝗶𝗹𝗹 𝗻𝗼𝘁 𝗯𝗲 𝗮𝗯𝗼𝘂𝘁 𝘄𝗵𝗲𝘁𝗵𝗲𝗿 𝗔𝗜 𝘄𝗮𝘀 𝘂𝘀𝗲𝗱. 𝗜𝘁 𝘄𝗶𝗹𝗹 𝗯𝗲 𝗮𝗯𝗼𝘂𝘁 𝘄𝗵𝗲𝘁𝗵𝗲𝗿 𝗮𝗻𝘆𝗼𝗻𝗲 𝗰𝗵𝗲𝗰𝗸𝗲𝗱 𝘄𝗵𝗮𝘁 𝗶𝘁 𝗼𝗽𝘁𝗶𝗺𝗶𝘇𝗲𝗱 𝗳𝗼𝗿. 🔗 in comments

AI is new and shiny. Employment law is not. Mobley v. Workday proves the point. The court held that employers don't get to outsource liability just because they've outsourced the tool to an AI vendor. The plaintiffs, a nationwide class of job applicants over the age of 40, allege that employers' use of Workday’s AI-driven screening tools discriminates on the basis of age. Whether those claims ultimately stick is a question for another day. But the legal framework governing them is old, settled, and very familiar. Discrimination is discrimination—whether it's carried out by a hiring manager, a spreadsheet, or an outsourced algorithm. What would have been surprising is the opposite outcome—if the court had said, "Not your problem, employer, your vendor did it." That's not how employment law works. It never has been. If your hiring process produces a disparate impact, you own it. Full stop. This case—and others like it percolating through the courts—should recalibrate how employers think about HR tech. AI doesn't create new legal obligations. It just exposes how seriously you're taking the ones that already exist. So what should you be doing now? Start with your contracts. If you're relying on a vendor's AI to source, screen, or rank candidates, you need to understand exactly how liability is allocated. Who is indemnifying whom? For what claims? With what caps and carveouts? "Trust us" is not a risk mitigation strategy. Next, build audit rights into those agreements—and use them. You should have the contractual ability to test your vendor's tools for disparate impact and to obtain meaningful information about how those tools function. If you can't evaluate it, you shouldn't be using it. Also, don't treat AI as a black box. You don't need to code it, but you do need to understand how it's trained, what data it relies on, and where bias might creep in. Speed and efficiency are great. Not at the expense of compliance. Finally, own the outcomes. If a tool flags—or filters out—candidates, that's your hiring decision. Regulators and courts aren't going to draw a distinction between "human" and "machine-assisted" discrimination. AI may be the new frontier. The rules governing it are not. Ignore that at your peril.