🎉The 2026 Radar is live! Interested in contributing? Get into contact with us!
highlighted
to the right, which point to the interconnectedness of many use cases.Knowledge Management lawyers and antitrust/competition lawyers within a large German law firm are using the tools. The firm also operates a dedicated legal tech venture that develops client-facing solutions and handles large-scale investigations.
The primary goals are knowledge management, automating administrative workflows, and assisting with business development. Specific use cases include generating internal newsletters on case law updates (which reduced the user's manual drafting time from 12 hours to 1 hour), automatically structuring event/webinar invitations from emails, and drafting legal publications. They are also experimenting with custom AI agents to act as "junior associates" for first-draft legal analysis using publicly available EU antitrust guidelines.
The firm officially provides Microsoft Copilot and an internal proprietary GPT-wrapper. However, to bypass the limitations of general-purpose tools, the user builds Custom GPTs via ChatGPT (utilizing the Code Interpreter function). They also use Gemini and Claude for specific tasks. To ensure highly precise citations and prevent probabilistic paraphrasing, public legal texts are heavily pre-processed and structured into JSON formats (defining articles, paragraphs, and sentences) before being fed into the models.
Production
Interview conducted in April 2026, led by sebis researchers. The interviewee is a Knowledge Management Lawyer specializing in antitrust law, working at the organization.
Lawyers across various decentralized departments, specifically in Real Estate, Tax, and M&A, as well as a Digital Transformation Officer who coordinates IT and digital processes. Tool adoption is largely bottom-up, driven by individual or departmental needs rather than a top-down, firm-wide mandate.
The firm uses AI primarily to process, review, and structure large volumes of legal documents efficiently, particularly for M&A due diligence and complex asset management. Specific use cases include preliminary legal research, summarizing and structuring historical dispute documents, quickly analyzing foreign legal texts (e.g., US Trust deeds) for German tax implications, and automatically generating corporate structure charts from text or handwritten notes. AI is also used for broader organizational tasks.
They utilize a diverse, continually evolving stack of both legal-specific and general-purpose tools. For legal research, they use publisher-integrated AI assistants like Otto Schmidt Answers, BeckChat, and Juris KI. Legora is heavily utilized for document review and data preparation, while Noxtura is currently in a testing phase. Jigsaw AI is employed to generate PowerPoint organizational charts. Additionally, they use general off-the-shelf LLMs like ChatGPT—having found that domain-specific tax models do not consistently outperform them—and have fully rolled out Microsoft Copilot for general administrative and organizational workflows.
Production
Interview conducted in April 2026, led by sebis researchers. The interviewees are a Digital Transformation Officer and a practicing Lawyer specializing in Real Estate and Tax, working at the organization.
The tools are developed for and used by lawyers in both large law firms (Big Law) and in-house legal departments of major corporations. Users range from highly specialized technical legal engineers and innovation teams to individual practitioners across various fields, including employment law, insolvency, and criminal law.
The provider offers two main solutions to address different legal needs. The first is a productivity tool for individual lawyers to assist with document-heavy tasks like summarization, comparison, drafting, and legal research. The second is a low-code/no-code automation platform that allows firms to transform manual processes into digital applications without deep coding knowledge. Key use cases include automated intake, self-service tools for clients (e.g., "Dawn Raid" apps), NDA automation, and streamlining repetitive administrative "office work" like client communication and file management.
The provider utilizes Generative AI (GenAI) and Large Language Models (LLMs), primarily integrating established American models due to their superior performance. They employ a mix of technologies, including semantic search, Retrieval-Augmented Generation (RAG) grounded in specialized legal databases (e.g., Otto Schmidt), and custom-trained anonymization techniques to handle sensitive data. The platform also features "Poka-Yoke" (error-proofing) mechanisms, such as visual validation markers (green checkmarks) that verify AI-generated citations against actual case law and statutes.
Production
Interview conducted in April 2026, led by sebis researchers. The interviewee is the CEO/founder of a legal technology company providing AI-based automation and productivity tools.
Legal professionals in the IT law and digitalization practice group at a large international law firm. The users focus on IT law, data protection, and EU digital strategy compliance (e.g., AI Act).
The tools are used to act as a "sparring partner" for legal brainstorming and to streamline daily workflows. Specific use cases include initial legal research, drafting and adjusting the tone of emails and pleadings, and restructuring standard contracts (e.g., works council agreements). A major solved problem is translation; AI is now used to provide "convenience translations" for international clients, which are faster and cheaper, completely replacing the firm's former internal translation department. The tools are also used for creating presentations and daily organizational tasks (e.g., email and to-do list management).
They use a combination of legal-specific databases and general-purpose large language models. Legal tools include Juris KI and publisher-specific bots (e.g., Beck's "Frag den Grüneberg"). For drafting and structural tasks, they rely heavily on Claude and Google Gemini. Because the firm does not have specific enterprise data agreements for these public LLMs, all inputs must be strictly anonymized manually. Microsoft Copilot is used within a protected environment, primarily for organizational tasks. Specialized legal AI tools like Harvey are not currently available due to pending global committee strategy decisions.
Production
Interview conducted in April 2026, led by sebis researchers. The interviewee is an IT law professional and associate working at the organization.
Lawyers across the firm globally and in Germany, particularly within data-heavy practice groups such as Litigation, Dispute Resolution, and Corporate/M&A. Adoption varies, with approximately 20% being heavy users, 30% occasional users, and the rest either testing sporadically or refusing to use the tools.
The firm uses AI primarily to handle massive volumes of unstructured data. Key use cases include extracting facts to build case chronologies in litigation, document review, and due diligence in corporate transactions. AI is also used for translation, general text correction (e.g., finding formatting errors or incomplete sentences), and as a brainstorming partner. The firm actively avoids using AI for drafting standard contracts, as their extensively maintained, automated internal templates yield superior and more consistent results.
They use a wide mix of generative AI and traditional deep learning tools. Generative AI tools include a custom ChatGPT environment hosted in Microsoft Azure, Microsoft Copilot, and Harvey. Specialized tools include DeepL for translation, Wechsler for fact extraction and chronologies, Orbital Witness for real estate, and Vlex for case law research in the US. They also rely heavily on trained deep learning systems like Kira and Relativity for document analysis, which currently outperform non-trained GenAI models in specific extraction tasks.
Production
Interview conducted in April 2026, led by sebis researchers. The interviewee is an IT and legal tech professional working at the organization.
AI tools are used by over 60% of the firm's personnel, including lawyers, associates, business professionals, and legal trainees (Referendare). While available firm-wide, usage and pilot programs are heavily focused on four key standardizable practice groups: Corporate, M&A/Real Estate M&A, Conflict Resolution (Litigation), and Employment Law.
The firm focuses on three main pillars for AI application: text creation/drafting, data analysis (such as reviewing large volumes of documents for due diligence), and legal research. Translations are also a common use case. A broader organizational goal is to establish AI-native workflows, adapt training programs so junior staff learn to verify AI outputs rather than perform manual searches, and prepare for future collaborative AI environments with clients.
The firm employs a deliberate "multi-lane strategy" rather than relying on a single one-size-fits-all provider. They provide base licenses for major legal AI platforms (Harvey or Legora) and Microsoft Copilot to their staff. They supplement these with specialized tools (like Noxthur) for specific research or data analysis tasks, and integrate smaller, domain-specific tools capable of querying specialized German and EU legal databases (e.g., Beck, Otto Schmidt).
Production
Interview conducted in April 2026, led by sebis researchers. The interviewees are legal innovation and transformation professionals working at the organization.
M&A lawyers and other legal professionals across multiple jurisdictions (EMEA, Americas, Asia) at a large international law firm. The global evaluation and rollout of tools are coordinated centrally by a dedicated innovation team, overseen by a Chief Innovation Officer, with localized training groups supporting adoption.
The firm aims to process massive volumes of documents in virtual data rooms that are no longer feasible to review manually. AI is used for due diligence analysis, generating contract markups and issue lists, and drafting smaller contracts and legal memos. Additionally, AI serves as a quality assurance tool to review contracts for clarity and inconsistencies, and to help structure legal arguments.
They primarily use a specialized legal AI platform called "Ligora" (partially powered by Claude) which is integrated directly into Microsoft Word. General-purpose models like ChatGPT (referred to internally via a system called Atlas) are also utilized, though with caution. Other tools tested or in use include Microsoft Copilot, Harvey, and specialized databases like Beck next*tour for German law research. The firm is highly interested in moving toward "Agentic AI" to integrate directly with internal document management systems like iManage and Outlook.
Production
Interview conducted in April 2026, led by sebis researchers. The interviewee is an M&A lawyer and Lead Innovation Partner for EMEA at the organization.
Lawyers at a large German law firm, particularly those specializing in M&A and finance. An internal AI task force meets monthly to coordinate the firm's AI strategy, evaluate tools, and drive adoption. Tool procurement is decided by the partnership. While usage is becoming widespread, adoption levels vary, ranging from skeptical older partners to dedicated "power users."
The primary goals are time savings, structuring complex thoughts, and general convenience. In M&A and finance, AI is used to analyze large data rooms—such as extracting "Change of Control" clauses from hundreds of documents simultaneously. It is also used to draft initial contracts, check existing drafts for logic and systematic errors, summarize large fact patterns into bullet points, and generate executive summaries based on due diligence reports. Additionally, the tools are used to extract relevant passages from legal literature for writing legal commentaries.
The firm relies heavily on Harvey (specifically utilizing its "Vault" feature for analyzing large document batches) and Microsoft Copilot (used more like a chat assistant for quick queries or brainstorming). Both tools are secured via enterprise agreements to ensure client data is not used for model training. ChatGPT was used early on but was abandoned to avoid the manual effort of anonymizing data. The firm tested Beck NoxxTura but found the AI capabilities lacking, and they are preparing to pilot Google Gemini soon.
Production
Interview conducted in April 2026, led by sebis researchers (TU Munich). The interviewee is an M&A lawyer and active AI power user, working at the organization.
Lawyers specializing in Intellectual Property (IP), Trademark, and Competition Law at a large law firm. The users include partners, associates, and legal assistants (WissMits), supported by an internal legal tech team and dedicated software developers.
The firm aims to enhance the quality and depth of legal work while increasing efficiency in research and drafting. Key use cases include factual research to explain technical products in trademark similarity cases, preparing for client meetings based on vague email descriptions, and analyzing large volumes of court decisions (e.g., EuG rulings). It is also used to summarize lengthy legal chapters into executive summaries, refine the linguistic precision of arguments, and automate administrative tasks such as meeting protocols, email searches, and converting Word documents into presentation slides.
The firm utilizes a diverse stack of general and specialized tools. General models include Gemini (specifically the Deep Research feature) and ChatGPT. Microsoft Copilot is deeply integrated into the Office 365 environment (Teams, Word, Outlook, PowerPoint). Specialized legal research is conducted via Beck-Nox Tour, and notebookLM is used for analyzing public datasets. Crucially, the firm uses "CRAG" (Client Refined AI Generator), an internally developed, secure AI tool built on OpenAI APIs hosted in a private cloud to ensure data protection and professional secrecy.
Production
Interview conducted in April 2026, led by sebis researchers. The interviewee is a lawyer and legal tech expert working at the organization.
Experienced M&A lawyers, particularly those specializing in complex, cross-border private equity and insurance transactions, at a large international law firm with over 2,500 employees. Junior associates also use the tools for foundational groundwork and initial document drafting.
The primary goal is improving the quality and precision of legal products rather than solely focusing on efficiency or reducing billable time. Key use cases include refining the idiomatic flow of English emails, acting as a sparring partner for complex contract drafting, summarizing massive document sets (e.g., thousands of pages for regulatory reviews), drafting project pitches, and comparing document drafts against historical precedents like scoping memos.
The firm uses enterprise versions of large language models (Claude, ChatGPT, Gemini) hosted within a proprietary, secure browser-based environment to handle confidential client data safely. NotebookLM is used for summarizing large volumes of documents. Perplexity is utilized for public market research, and a specialized chatbot integrated into the beck-online database is tested for legal research. The firm also utilizes custom-developed internal AI tools for specific workflows, such as generating document markups.
Production
Interview conducted in April 2026, led by sebis researchers. The interviewee is an experienced M&A lawyer specializing in insurance transactions, working at the organization.
The primary users are courts and administrative bodies, particularly in Switzerland, that need to anonymize legal decisions for publication. The tool is also used by national authorities and public institutions (e.g., BfArM) that process documents under freedom of information principles. Smaller to mid-sized law firms show interest but are often constrained by budget and support requirements.
The provider addresses the resource-intensive and complex task of anonymizing Word and PDF documents. The goal is to ensure the reliable protection of personal data while maintaining the readability of legal texts through the use of placeholders instead of simple redaction (blacking out). Key challenges solved include reducing the manual effort of first-pass document reviews and avoiding technical pitfalls in PDFs where sensitive data might remain hidden beneath redaction layers.
The software utilizes specialized Named Entity Recognition (NER) models to identify candidates for anonymization, such as persons, organizations, and locations. Instead of large, cloud-based LLMs, the provider uses smaller, fine-tuned models that are deployed on-premise at the client's site for data security. The system combines these AI models with rule-based approaches like regular expressions (Regex) and pattern recognition. It is delivered as a Word Add-in that operates through a browser-based interface (Edge) within the application.
Production
An M&A specialist with over 20 years of experience, who recently transitioned from a Managing Partner role at a top 10 global law firm to a solo advisory practice. In the context of the global firm, users included partners, associates, and IT-savvy professional support staff who managed the technical processes and workflows.
The primary goal is to support complex M&A transactions and process massive volumes of data efficiently. Use cases include conducting legal due diligence by screening data rooms for specific conditions (e.g., non-compete clauses), summarizing extensive email threads within internal document management systems, and assisting in drafting standard contract elements like options agreements. Furthermore, AI is utilized for global knowledge management and CRM, allowing lawyers to query internal databases to identify past deal structures or locate specific industry expertise across the firm's global network.
The core technology utilized is a proprietary, internal AI system developed by the global firm, which automatically masks and anonymizes confidential client data before processing it through underlying external models. General-purpose models like ChatGPT are used occasionally for brainstorming or drafting simple explanations. Domain-specific AI tools, such as the assistant integrated into beck-online, are consulted for pure legal research. The interviewee also notes market awareness of contract analysis tools like SAP's AI features for benchmarking market standards.
Production
Interview led by sebis researchers. The interviewee is an M&A specialist and former Managing Partner of a global law firm, currently operating in a solo advisory capacity.
Legal tech professionals, lawyers (ranging from associates to partners), young talents, and personal assistants (PAs) at a large German law firm. Specialized departments, such as the IT and investigations teams, also utilize these tools for specific mandates and large-scale document analysis.
The firm uses AI to automate and streamline daily legal and administrative tasks. Use cases include drafting and refining emails, summarizing lengthy communication threads, and generating contracts from scratch or via standard playbooks (e.g., filling investment agreements from cap tables). In the context of internal investigations and e-discovery, AI is used to process terabytes of data to extract specific communication events and build factual chronologies. They also utilize AI to automatically generate visual timelines for corporate transactions and are exploring ways to automate threshold notifications in antitrust law.
They primarily use Harvey (including its Word/Outlook add-ins and Harvey Vault) as their generalist AI tool for drafting, document analysis, and daily tasks. For specialized legal research involving commentaries, they are testing Beck nocs Tour, though it is currently viewed as linguistically weaker than global LLMs. They also use an internally developed IT tool for investigations and rely on AI-enhanced software like Structure Flow for data visualization. Other enterprise models, such as ChatGPT Enterprise and Anthropic's Claude, were previously evaluated in comparative tests.
Production
Interview conducted in April 2026, led by sebis researchers. The interviewee is a lawyer specializing in legal technology, working at the organization.
Lawyers across various levels of seniority, from young associates to partners, are utilizing legal AI tools. The users span multiple practice areas, including litigation, corporate law, and intellectual property (IP). A dedicated internal "Taskforce KI" actively monitors the market, evaluates tools, and guides adoption. Legal assistants also use general AI tools for administrative tasks and translation, though they do not have access to the specialized legal LLMs.
The firm uses AI to accelerate initial brainstorming, legal research, and document drafting (e.g., generating first drafts of GmbH articles of association or specific clauses). AI is also heavily used for document analysis, such as processing large volumes of annexes in court cases or conducting due diligence. Additional use cases include translating texts, correcting style, summarizing documents, creating structured visual outputs like timelines or tables, and even generating Python scripts to automate routine office workflows (e.g., filtering Outlook emails).
The firm employs a variety of cloud-based language models. Harvey is their primary tool for legal drafting and analysis because it meets their strict professional compliance requirements. ChatGPT is used for general, technical, and non-legal brainstorming, but strictly with anonymized data due to OpenAI's refusal to sign customized compliance declarations. For targeted legal research, they use database-linked tools like Juris KI Suite and beck-Noxx-Tool. Microsoft Copilot and DeepL Pro are used for office integration and translations. They are also preparing to introduce Google Gemini (specifically NotebookLM) because it satisfies their regulatory and confidentiality standards.
Production
Interview conducted in April 2026, led by sebis researchers. The interviewees are a senior litigator and a junior associate, both working at the organization.
In-house legal counsel advising the IT department of a large German multinational corporation. The users act as commercial and IT generalists, handling a high volume of daily requests, contract negotiations, and internal legal coordination.
The primary goal is to increase efficiency in administrative and back-office tasks, such as sorting emails, finding documents, and drafting short communications. AI is highly valued as a feedback tool to simplify complex legal language into accessible internal corporate guidelines (e.g., directives on AI usage) for non-lawyers. While they are exploring AI for routine contract review (like NDAs) and tracking global legislative changes, they note that their core business—strategic contract negotiation and risk assessment—cannot yet be effectively automated.
The legal team uses Microsoft Copilot for daily back-office support and Azure OpenAI for general experimentation. For domain-specific tasks, they have tested and are acquiring licenses for Beck NexTur, primarily for legal research and summarization. They also evaluate various third-party contract review tools and occasionally experiment with building internal prototypes, though they rely heavily on pre-existing, highly customized corporate templates rather than AI drafting.
Production
Interview conducted in April 2026, led by sebis researchers. The interviewees are two in-house legal professionals working at the organization.
Legal technology consultants are using these tools to advise and enable in-house legal teams at corporate clients. The focus is on legal transformation, Legal Ops, and the implementation of Contract Lifecycle Management (CLM) systems for these corporate legal departments.
The primary goal is to digitalize and automate the contract lifecycle, including AI-assisted redlining, drafting, and extracting metadata from legacy contracts to populate CLM systems. Additionally, AI is heavily utilized for non-core legal and administrative tasks, such as chronologically summarizing massive, complex case files and verifying rule applications in documents (e.g., invoice audits). These administrative tasks yield the highest time savings, often reducing workload by up to 90% compared to strictly legal tasks.
They use AI capabilities integrated directly into specialized CLM platforms (such as Lia and Sirion). For generative and analytical tasks, they utilize a proprietary internal ChatGPT derivative, Microsoft Copilot, and Harvey, with which they hold an exclusive partnership. For deep legal research requiring specific German case law, they rely on specialized local databases like Beck Noctua, as general AI models lack access to this paywalled data.
Production
Interview conducted in April 2026, led by sebis researchers. The interviewee is a legal technology consultant advising corporate in-house legal teams.
Legal professionals, particularly within the patent law practice group, including knowledge lawyers, partners, and scientific assistants. Usage spans across demographics but is exceptionally high (around 90%) among younger employees.
The primary goals include managing internal case data, tracking deadlines, legal research, and drafting correspondence to overcome the "blank page" syndrome. A specialized use case in patent law involves summarizing and extracting knowledge from Unified Patent Court (UPC) decisions to build internal databases. Additionally, AI is used to quickly explain complex technical and scientific concepts (e.g., chemical formulas, telecommunications standards) required for patent litigation.
The firm relies heavily on Microsoft Copilot integrated into standard office applications (Word, Outlook, Power BI). They also use an internal proprietary AI assistant based on OpenAI's GPT models, developed via a joint venture. Legal publisher tools like Beck-online's NoxxTua and Beck Chat are currently in testing phases. For non-confidential, public data tasks, power users strongly prefer Anthropic's Claude models (specifically Opus) due to their advanced reasoning and tool-use capabilities.
Production
Interview conducted in April 2026, led by sebis researchers. The interviewee is a Patent Knowledge Lawyer, working at the organization.
Lawyers across various practice groups, including Tax, Private Clients, Corporate, and Regulatory, are using the tools. Partners are also increasingly driving adoption due to market awareness. A dedicated internal Innovation Lab consisting of 14 experts acts as a service provider to build, customize, and support these tools for both tech-savvy lawyers and broader firm use.
The most prominent use case is data extraction from legal and financial documents, such as tax assessments and compliance records. AI is also used for drafting support, particularly for emails, initial outlines for pleadings, and non-core legal tasks like preparing presentations or webinars. Furthermore, they are actively testing AI for document-heavy due diligence processes and using it for legal research, with future plans to apply AI to CRM and general law firm administration tasks.
They use self-hosted large language models (including GPT and LLaMa) accessed via an internal chatbot. The Innovation Lab develops complex expert tools using prompt chaining to produce structured data outputs (e.g., JSON). They also utilize AI modules natively integrated into standard legal databases like beck-online and Juris. Additionally, they have access to Microsoft Copilot and are currently running proofs of concept (POCs) for commercial legal AI tools such as Harvey, Legora, and Libra.
Production
Interview conducted in April 2026, led by sebis researchers. The interviewee is a legal professional working in the Innovation Lab of the organization, with a background in litigation.
Law firms (ranging from small practices to large commercial firms), corporate legal departments, and insolvency administrators. The end-users include lawyers, insolvency specialists, and back-office/secretarial staff handling administrative processes.
The provider aims to optimize both core legal work (contract drafting, review, and analysis) and repetitive administrative tasks (billing, document filing, controlling). In insolvency law, AI is used for text recognition, document summarization, and claims registration. The overarching goal is to integrate "Embedded AI" into existing firm management and document management systems to act as a digital backbone, thereby improving efficiency and mitigating the impacts of skilled labor shortages.
They use a model-agnostic approach, integrating various large language models including OpenAI (via Microsoft Azure), Claude (via AWS Bedrock), and Google Gemini, ensuring compliance with GDPR and professional secrecy laws. Customers can sometimes choose the underlying model. Different models are strategically used for specific steps, such as using one model for document review and another for drafting. They also employ Retrieval-Augmented Generation (RAG) and automated secondary validation loops (e.g., link checkers and secondary AI verification) to minimize hallucinations.
Production
Interview conducted in April 2026, led by sebis researchers. The interviewee is a representative and in-house counsel (Syndikus) at a Legal Tech software provider.
Lawyers specializing in Venture Capital (VC) and M&A transactions at a German law firm are the primary users. Usage includes partners and associates, supported by an internal Tech & Automation team and in-house developers who build custom MVPs and prototypes.
The main goal is to automate the due diligence (DD) process, particularly for seed and pre-seed startups where review criteria are highly repetitive. They use AI to identify "red flags" or dealbreakers, perform semi-automated document reviews, and structure large volumes of data. Other use cases include checking the "Chain of Title" for capitalization history, summarizing negotiation notes into contract changes via Word add-ins, and analyzing market standards for VC investments through clause databases.
They primarily use Legora, a specialized legal automation technology, for tabular document reviews and due diligence. They also utilize Microsoft Copilot for broader tasks and have developed internal tools for managing document requests from founders. For software development and experimentation, tools like Claude are monitored, though data security remains a concern. The firm favors a "build" approach for specific processes while using cloud-based platforms for standard legal tech tasks.
Production
Interview conducted in April 2026, led by sebis researchers. The interviewee is a lawyer specializing in VC and M&A, working at the organization.
Lawyers and legal professionals at a large international law firm are the primary users. This includes senior partners, associates, and specialized tech transaction teams. Implementation is supported by a dedicated Chief Innovation Officer and an innovation team, alongside a global Tech Committee that manages the onboarding process for the firm's legal tech stack.
The firm aims to automate repetitive drafting tasks, improve document quality through automated proofreading (identifying inconsistencies and errors), and streamline the negotiation process. AI is used to compare incoming vendor terms against internal playbooks to identify acceptable fallback positions. Additionally, the tools help manage email correspondence related to contract updates and assist in the drafting of standard agreements, such as SaaS contracts, to save time and enhance consistency.
The firm utilizes a multi-layered approach including a proprietary chatbot hosted within their own Azure Cloud environment to ensure data security. They use Legora (integrated as a Word plugin and Outlook extension) for drafting and proofreading, and Vincent AI (vLex) for accessing public legal databases. The firm also employs various large language models (LLMs) via APIs and has explored Microsoft Copilot and Harvey. Strategy focuses on "Buy & Build," ensuring tools are integrated directly into existing workflows.
Production
Interview conducted in March 2026, led by sebis researchers. The interviewee is an experienced lawyer and Of Counsel at the organization, with a long history in tech transactions and innovation committees.
A senior lawyer with nearly 30 years of experience, specializing in commercial, corporate, and insolvency-related law, is using legal AI. The firm’s internal IT department and a dedicated working group—including younger associates interested in legal tech—manage the evaluation and procurement of these tools. Usage intensity varies across the firm, with litigators and advisory teams using it more frequently than procedural departments.
The primary goals are to streamline the entry into legal research, summarize complex case facts, and generate initial drafts for legal statements or memos. AI is used as a "digital assistant" to provide a starting point for topics where the user is not the primary expert (e.g., labor law). Additionally, the firm aims to digitize administrative workflows, such as file creation and office management, to potentially reduce the need for administrative staff and increase overall efficiency.
The practitioner uses specialized German legal AI tools including Beck Noxtua, Juris (integrated AI features), and Beck Online Chat. These tools are used for document uploads, summarization, and research assistance. For private use and international or public law queries (e.g., international law), general-purpose models like ChatGPT are utilized. The workflow often involves using AI for an initial summary or "brainstorming" of points to consider before moving to traditional legal commentaries.
Research
Interview conducted in February 2026, led by sebis researchers. The interviewee is a senior lawyer and partner specializing in commercial and insolvency law.
The primary users are lawyers and legal engineers within a German law firm, as well as legal departments of large organizations (including DAX companies) that receive consulting on AI integration. The interviewee is a legal professional and legal engineer who develops custom AI solutions for various legal fields such as IT law, data protection, and procurement law.
The goals include massive efficiency gains in document-heavy processes. Key use cases involve contract review (e.g., verifying existing portfolios against new regulations like DORA), due diligence, and legal research in both public sources and specialized databases. Specific focus is placed on automating routine tasks like checking service charge statements for real estate portfolios and performing automated civil law assessments (BGB) to increase "Access to Justice" for small-scale legal claims.
They use a custom-built platform called "Digital Gateway Gen AI," based on Microsoft Azure, which integrates various LLMs including GPT-4, GPT-5.2 (experimental), Gemini, and Claude. They employ an "ecosystem of personas" (pre-configured agents) and are developing agentic systems where one orchestrator LLM controls other agents to perform complex, multi-step workflows. Microsoft Copilot is also used horizontally for standard office tasks, while specialized tools like Legora and Harvey are tested for vertical integration.
Production
Interview conducted in April 2026, led by sebis researchers. The interviewee is a practicing lawyer and legal engineer with nine years of professional experience, working at the organization.
The tools are developed for a wide range of legal professionals, including lawyers in small family-owned firms, specialized boutique firms (e.g., in corporate or labor law), and in-house legal and procurement departments of large corporations. The provider also collaborates with major legal publishers (e.g., Lex Verlag) and integrates with standard office software like Microsoft Word and SharePoint.
The provider aims to reduce the time and effort required to create and finalize legal documents. Key use cases include automating the drafting of standard contracts (like employment agreements), providing "negotiation toolkits" for lawyers, and helping users who lack deep legal expertise (such as procurement staff) to work with standardized templates. The goal is to bridge the gap between "black box" AI and expert human knowledge by providing calibrated, non-opaque solutions.
The provider uses a multi-model approach, integrating various Large Language Models (LLMs) via APIs, including OpenAI (GPT-4), Google, and Mistral. They employ an internal benchmarking system to select the best-performing model for specific functions. For data-sensitive environments, they offer deployments on dedicated servers (e.g., via AWS or Microsoft Azure) to ensure compliance with professional secrecy and data protection standards. They also utilize "Transformer" models and explore hybrid approaches for template automation.
Production
Interview conducted in April 2026, led by sebis researchers.
The primary users are legal professionals within the Corporate Governance department, which includes legal counsel, compliance officers, internal auditors, and risk management staff. The department is a frontrunner in AI adoption within the company, with usage extending from associates to the General Counsel.
The organization aims to improve efficiency and resource allocation by automating low-complexity, repetitive tasks. Specific goals include summarizing large volumes of documents for internal audits, providing self-service legal advice via a policy chatbot, and conducting regional-specific legal research (e.g., a "Chinese Paralegal" assistant). They also work on automating the translation of standard operating procedures into visual process diagrams and performing regulatory horizon scanning to identify changes in legislation.
The firm is heavily integrated into the Google ecosystem, primarily using Gemini as their central productivity and generative AI tool. They utilize "Custom Gems" for tailored tasks and have experimented with Deep Research Models for niche legal questions. Additionally, they have recently integrated Noctua, a specialized legal AI tool, for tasks requiring high precision and data segregation from the internet. They also use automated coding assistants to create App Scripts for workflow automation within spreadsheets.
Production
Interview conducted in April 2026, led by sebis researchers. The interviewee is a legal professional responsible for AI rollout and digital transformation within the organization.
A judge at a German civil court is using AI tools to assist with judicial tasks. While the judiciary is generally considered a "AI-free zone" due to a lack of specialized access, individual judges are beginning to test official pilots and use general-purpose models to support their work.
The judge aims to improve the quality and depth of judicial decisions rather than simply increasing speed. Key use cases include researching case law, summarizing case files, creating timelines of events from legal briefs, and identifying conflicting versus undisputed party claims. AI is also used as a "quality controller" to ensure all arguments have been addressed and to brainstorm legal interpretations or potential questions for witnesses and experts.
The primary tools used are Beck-Noxx (a specialized legal AI by Beck-Verlag currently in a pilot phase) and Claude (specifically the Opus 4.6 model). Because of strict data protection requirements, an anonymization tool developed by the University of Erlangen-Nuremberg (FAU) is used to redact personal data before documents are uploaded to AI interfaces. The judge also utilizes "Deep Research" functions for abstract legal problems and experiments with prompting to break down complex legal questions.
Research
Interview conducted in February 2026, led by sebis researchers. The interviewee is a judge working at a German civil court.
About Us | Contact Us | © sebis 2026 | sebis Homepage | Imprint