Artificial intelligence and legislative changes mean that organisations will face an increasing volume of lengthy claims. Too often, expensive lawyers advise a commercial settlement, because fighting a claim is more expensive than paying someone off, even when they do not have a good case.
DefendThis is an AI-powered legal defence platform created by ranked barristers. It helps businesses respond to employment, civil and commercial claims more accurately and more efficiently. You upload your documents. Leonidas reads them, asks targeted questions of the people with knowledge of the facts, and produces a professionally drafted defence.
NHS trusts, local authorities, police forces and universities (to name but a few) face a growing volume of civil and commercial claims. In employment tribunals alone, 42,000 single claims were received in 2024/25. The burden is increasing: the Employment Rights Act 2025 reduces qualifying periods for key rights and widens the scope of claims that employers must answer. The British Chambers of Commerce puts the average employer cost of defending one at £8,500. For many cases it is much more. For a public body handling thirty or forty claims a year, that is a quarter of a million pounds before a hearing is listed.
DefendThis takes the core task of responding to civil and commercial claims and runs it through AI. It does not stop at a single document. The platform produces a Grounds of Resistance or a letter of response, witness statements, a counter schedule of loss, a list of issues, and draft instructions to counsel.
The AI cross-references each witness's account against the uploaded documents. Where the account conflicts with the paperwork, it presses for an explanation. Where evidence is missing, it says so. The output is a better set of case documents than most solicitors produce from a first conference, and it takes hours rather than weeks.
Defending a claim in a large organisation is slow. HR or in-house legal chases managers for their account. Managers are busy, spread across sites, unclear about what is needed. Emails go back and forth. Phone calls are scheduled and rescheduled. Weeks pass. DefendThis replaces that process.
The preparation work that currently absorbs your in-house team's time is done by the platform. Your lawyers receive a draft defence or letter of response, witness statements, a counter schedule, a list of issues, draft instructions to counsel, and an indexed document bundle. The documents are more accurate than those produced by the conventional process, because every factual assertion has been tested against the uploaded evidence before the draft is produced.
Your in-house team reviews, amends, and files. The hours they currently spend chasing witnesses, collating documents, and drafting from incomplete notes are returned to them for strategic work.
Beyond the standard response workflow, DefendThis builds bespoke products for public sector clients. Witness statement generators that take instructions from individual witnesses and produce drafted statements ready for legal review. Counter schedule of loss tools that calculate and present the respondent's position on remedy. Case management systems that track every claim across your organisation from receipt to disposal.
Each product uses the same AI and the same barrister-designed protocols. They are built to your specification and sit within the DefendThis dashboard your team already uses.
Large organisations face a steady flow of claims. Employment tribunal claims, commercial disputes over contracts and non-payment, civil claims from customers, suppliers and neighbours. Each one triggers the same machinery. Instruct external solicitors. Gather the facts from the people inside the business. Draft the response. File. Repeat. The cost is significant and largely uncontrolled. For organisations defending dozens of claims a year across different areas of law, the annual spend runs well into six figures. DefendThis prepares the first draft internally, across every type of claim, faster and at lower cost than any external firm.
DefendThis is built as a bespoke platform for each corporate client. The AI is loaded with your organisation's own policies, contracts, and internal procedures, across every area where claims arise. Your employment policies, your supplier contracts, your terms of business, your consumer-facing terms and conditions. When the system questions a manager, it cross-references their account against your documents, not generic templates.
The documents produced are more accurate than those from the conventional process. Every factual assertion is tested against the uploaded evidence before a draft is generated. Inconsistencies are identified and resolved at the fact-gathering stage, not discovered by counsel the week before the hearing.
Your in-house team receives a complete case file, whatever the claim type. For a tribunal claim, that is a Grounds of Resistance, witness statements, a counter schedule of loss, a list of issues, and draft instructions to counsel. For a civil or commercial claim, a CPR-compliant defence, witness statements, a schedule of disputed facts, and draft instructions. They review and refine rather than draft from scratch.
In most large organisations, defending a claim starts with chasing. In-house legal chases HR. HR chases the manager who made the decision. The manager is in a different office, or a different country, and does not know what information is needed. Weeks of emails produce an incomplete account. DefendThis replaces that.
The corporate platform includes a case management dashboard. Your in-house legal team or HR director sees every active claim across the organisation in one place: claim type, status, deadline, and the complete case file when it is ready.
Your in-house legal team creates cases and assigns them to the people who hold the facts. Each person receives a link and completes their part independently. The dashboard tracks who has responded and who has not.
The standard DefendThis platform produces a Grounds of Resistance or a defence to a civil or commercial claim. For corporate clients, we go further. We can build bespoke witness statement generators that take instructions from individual witnesses and produce drafted statements ready for review. We can build counter schedule of loss tools that calculate and present the respondent's position on remedy. We can build case management systems that track every claim across your organisation from receipt to disposal.
Each of these products uses the same AI, the same evidence base, and the same barrister-designed protocols. They are built to your specification and integrated into your DefendThis dashboard.
The corporate product is delivered as a bespoke build: your policies loaded, your procedures configured, your dashboard set up, your users created. That is a one-off cost. Thereafter, an annual licence fee covers unlimited access to the platform, ongoing AI updates, and support.
No per-claim charges, save for minimal token costs. No hourly billing. A fixed, predictable cost for claim response preparation across the organisation.
Maria works for an international audit firm. She brings a claim for failure to make reasonable adjustments based on her mental health conditions, together with a series of grievances. Follow the claim through the platform, step by step.
| Stage | Action | Output |
|---|---|---|
| 1 | £300 fee paid | Access to the platform |
| 2 | Upload the claim and supporting certificate | Claim analysis and deadline confirmation |
| 3 | Upload contracts, records, correspondence | Indexed bundle with gap analysis |
| 4 | AI-guided interview: structured questions, plain English | Full transcribed account |
| 5 | AI drafts the response | Grounds of Resistance or letter of response (draft) |
| 6 | Review and confirm each factual assertion | Client-verified response document |
| 7 | Partner solicitor reviews and files | Filed response |
| 8 | Ongoing case support as the matter progresses | Witness statements, correspondence |
The interview follows protocols written by ranked employment barristers. It takes the factual chronology, identifies the decision-makers, establishes the reason for each disputed decision, and then tests. Where your account conflicts with the documents, it does not move on. It asks why. Where a gap in the evidence exists, it asks what happened.
This replicates how a barrister prepares a witness before tribunal. The interview is transcribed and forms the foundation of the drafted response. You confirm each factual assertion individually before the document is finalised.
Most AI legal companies are run by non-lawyers, trying to sell complicated generalist products to law firms. DefendThis is different. It has been founded by three senior barristers, who have encoded their expertise about how courts and tribunals operate into the model. DefendThis provides organisations with a series of simple tools that drastically reduce the amount it costs to defend claims.
Despite hundreds of AI tools being on the market, costs of fighting claims remain enormous. The reason is structural. Firms whose revenue depends on billable hours have little reason to adopt tools that shorten the time spent on a client matter. The AI products marketed at law firms are designed to fit around the billable hour model, not to displace it.
DefendThis is built for the people who pay those bills. It sits with the organisation defending the claim, not the firm representing it. The output is a complete first draft of a defence or witness statement, produced in hours rather than weeks, and at a fraction of the usual costs. Internal lawyers can then check the output, or send a complete set of documents to external firms, saving time and cost.
Insurers are dealing with a triple burden: higher claim volume, higher claim values, and higher claims management costs. Across motor, EL/PL, property damage and general casualty insurance, thousands of claims handlers and loss adjusters incur billions of pounds in the costs of processing and managing liability claims. Every insurance claim is as different as it is similar. Common patterns, but different facts. DefendThis cuts the time it takes to respond.
DefendThis is an AI system, developed by ranked barristers with specialisms in each claims area. It takes the materials of a claim, absorbs them, and works out how to respond.
It produces all the documents required to defend an insured liability claim. A response to a letter of claim, a CPR-compliant Defence, witness statements, counter schedules, Part 18 and Part 35 questions, and draft instructions to counsel.
The AI understands the details of the claim and works out what is needed to respond. It questions the witnesses and cross-references their accounts with uploaded documents. It presses for detail, demands explanations, and probes inconsistencies. Where evidence is missing it says so.
DefendThis produces in minutes what claims handlers and solicitors take weeks, and dozens of billable hours, to produce. With greater consistency, and according to master strategy instructions provided by the insurer client.
DefendThis is flexible enough to fit the requirements of any insurer or claims handler. At its core, it pulls together the process of gathering and processing evidence. It cuts out the phone calls, emails and endless chasing that bog down the claims management process.
Instead of spending their time, and your money, wading through a diffuse sea of papers, panel solicitors receive clear, organised, focused documents for review. The fact-finding has been done. Your lawyers can focus on what they do best. Analysis, strategy, and advice.
Every insurer is different. DefendThis is a flexible platform, and can be adapted to suit any organisation, without compromising its core function. Every kind of dispute dealt with by insurers can be accommodated. The same AI and the same barrister-designed rules and protocols work throughout, configured to your house style, your reserving philosophy, and your strategic priorities.
DefendThis is built by expert, experienced barristers who can see that the current system often fails businesses. The founders have spent their careers preparing witnesses, drafting defences, and watching clients spend thousands of pounds on work that a well-trained AI can do faster and more accurately. DefendThis exists because they decided to do something about it.
Barrister, Outer Temple Chambers. Called 2015. Ten years' experience in employment law and commercial litigation. Ranked as a Leading Junior in Legal 500 (Tier 4, Employment) and in Chambers and Partners.
Gus designed the DefendThis platform, the AI instruction set, and the interview protocols. He maintains a practice at the Bar while building the company.
Barrister, 12 King's Bench Walk. Called 2015. Head of the 12KBW employment team. Ranked as a Leading Junior in both Chambers and Partners and Legal 500 for employment law. David's practice spans employment, personal injury, and civil litigation. Contributing editor to Harvey on Industrial Relations and Employment Law, the leading practitioner textbook. Fellow of the Royal Statistical Society.
David is building the AI's protocols for legal reasoning, evidence testing, and drafting. He is building the panel of specialist barristers whose expertise is embedded in the system.
Barrister, Littleton Chambers. Ranked as a Leading Junior in both Chambers and Partners and the Legal 500 for employment law. Nominated for Employment Junior of the Year at the Legal 500 Bar Awards 2025. Littleton Chambers won Employment Set of the Year at the same awards.
Grahame is an expert in High Court business protection disputes, acting in high-value team move litigation and restrictive covenant injunctions for major corporates. His wider practice covers discrimination, whistleblowing, and complex cases involving questions of territorial jurisdiction.
It does not provide legal advice, any more than ChatGPT or Claude do. It produces draft defences and letters of response for regulated lawyers to review.
DefendThis does not conduct litigation, exercise rights of audience, or provide legal advice in the reserved sense under the Legal Services Act 2007. It prepares documents, organises evidence, and produces written materials based on the information the client provides. The documents it produces are drafts. They are marked as drafts. They require review by a regulated lawyer before they are filed with any court or tribunal.
This is the same position as any other AI tool that produces written output. The difference is that DefendThis is built by practising barristers who have designed the system to produce output that a solicitor can review quickly, rather than output that requires substantial reworking.
For public sector bodies and large corporates, the review is conducted by in-house solicitors or the organisation's panel firm. DefendThis does the preparation. The regulated lawyer reviews it, gives advice where needed, and files. The time and cost savings are substantial because the lawyer is reviewing a near-complete file, not starting from scratch.
For small and medium enterprises without in-house legal, DefendThis partners with solicitors' firms to conduct the review and provide regulated legal advice at fixed fees. The solicitor receives the draft defence or letter of response, the indexed bundle, and the full record of the fact-finding process. Because the preparation is done, the review is faster and cheaper than it would be if the solicitor had taken the case from the outset.
The client receives regulated legal advice. The solicitor files with the tribunal. DefendThis handles the preparation that would otherwise have consumed dozens of hours of billable time.
DefendThis will handle sensitive legal documents: employment contracts, disciplinary records, medical reports, grievance correspondence. The platform is being designed so that DefendThis as a company will have no access to client case materials.
All uploaded documents and case materials will be encrypted at rest (AES-256) and in transit (TLS 1.2 or above). Each client's data will be segregated. DefendThis staff will not be able to access client case files. The AI will process the documents to produce the output, but no human at DefendThis will read them.
All case data will be automatically deleted either on a date the client selects, or 15 days after the client's last interaction with the case, whichever comes first. The client will be warned 48 hours before deletion and will be able to download their materials beforehand. No case data will be retained beyond the deletion window. No case data will be used for model training.
DefendThis will process personal data as a data processor on behalf of the client (the data controller). Processing will be conducted on a lawful basis, with clear privacy notices, defined retention periods, and documented data processing agreements. The client will control their data at all times. Subject access requests, deletion requests, and data portability requests will be supported through the platform.
All data will be stored in UK-based data centres. No case data will leave the United Kingdom.
DefendThis will pursue Cyber Essentials Plus certification, the UK government-backed standard required for suppliers handling sensitive or personal data in public sector contracts. Cyber Essentials Plus requires independent technical audit of five core security controls: firewalls, secure configuration, access control, malware protection, and patch management. It is a mandatory requirement for Ministry of Defence suppliers and for central government contracts involving personal data.
Beyond Cyber Essentials Plus, DefendThis intends to achieve ISO 27001 certification, the internationally recognised standard for information security management systems. ISO 27001 covers a broader range of controls including risk management, business continuity, and ongoing audit and improvement. It is widely regarded as the most rigorous information security standard available.
The documents produced by DefendThis will be created for the dominant purpose of defending legal proceedings. They are likely to attract litigation privilege. The client will own the documents. The privilege will belong to the client.
DefendThis produces two categories of document. Leonidas produces draft Grounds of Resistance (for employment tribunal claims) and CPR-compliant defences (for civil and commercial claims). Anita produces litigation-privileged witness proofs. All outputs are marked DRAFT. They are designed to be reviewed, amended, and filed by a qualified lawyer.
No. DefendThis replaces the fact-gathering and initial drafting stages of the litigation process. A qualified lawyer must review every document before it is filed with a court or tribunal. For organisations with in-house legal teams, the lawyer is the in-house solicitor or general counsel. For SMEs without in-house legal, DefendThis will recommend a partner solicitor who reviews and files at a fixed fee.
Each witness receives a secure link. They answer questions and upload documents from their own desk. The AI tests every answer against the uploaded paperwork. Where the witness's account conflicts with a document, it asks why. Where evidence is missing, it says so. The process is designed by barristers who have spent a combined 35 years litigating cases like the ones the system handles.
At launch, DefendThis will handle employment tribunal claims (unfair dismissal, discrimination, whistleblowing, wages), county court civil claims, and commercial disputes. The platform is designed to extend to further claim types as the product matures.
The traditional process of gathering facts, chasing witnesses, drafting and redrafting takes weeks. DefendThis compresses that into hours. Each witness typically completes their part in 15 to 30 minutes at their desk. The draft is available for review on the same day the last witness completes.
Nobody at DefendThis. The platform is designed so that DefendThis staff have no access to client case materials. All documents are encrypted at rest (AES-256) and in transit (TLS 1.2+). Each client's data is segregated. All data is stored in UK-based data centres. Case data is automatically deleted either on a date the client selects, or 15 days after the client's last interaction, whichever comes first.
No. No case data is used for model training. The AI models used by DefendThis are commercial frontier models accessed via API. Client data is processed to produce the output, then deleted according to the retention policy. It is never retained for training purposes.
The documents produced by DefendThis are created for the dominant purpose of defending legal proceedings. They are likely to attract litigation privilege. The privilege belongs to the client.
Three barristers with thirty-five years of combined experience at the Bar. The founders have experience across employment law, civil litigation, and commercial disputes. The AI operates using protocols designed by the founders, drawing on their experience of drafting defences, proofing witnesses, and advising on settlement.
DefendThis Limited is not a law firm and does not provide legal advice. It provides technology that assists with the preparation of legal documents. The draft documents are always reviewed by a qualified lawyer before filing. The founders are individually regulated by the Bar Standards Board.
Leonidas, the defence drafter, is available in prototype form. The AI is live. You can upload a claim document and work through the full process. Allow 15 to 20 minutes. The output is a working draft defence. We are looking for early conversations with organisations that want to test the platform.
Pricing has not been finalised. The model will vary by organisation type. For large organisations and the public sector, a subscription model is likely. For SMEs, a per-case fee bundled with the partner solicitor's fixed fee is under consideration. We are in early conversations and welcome discussions about pricing structures that work for different types of organisation.
You are the Head of Employment Law (a qualified solicitor) at a national chain of coffee shops. The manager of your Aberystwyth branch has recently dismissed their deputy manager for having a relationship with a member of staff. The deputy manager brings an unfair dismissal claim and makes various allegations of sex discrimination against the manager.
You are the in-house lawyer at a hotel chain. You receive a Letter of Claim for £30,000 from a couple whose wedding was held at one of your venues. They say the event was ruined by your staff. The food was late. The DJ was cancelled without notice. A member of staff was rude to the bride's mother. They want a full refund and damages.
Leonidas reads the claim, extracts every factual allegation, conducts a structured interview following protocols written by ranked barristers, and drafts a Grounds of Resistance or CPR compliant defence. The prototype is live and produces real output from a live AI model.
Leonidas takes a Particulars of Claim or ET1 and works through it paragraph by paragraph. It identifies background facts (parties, dates, terms) and core contested allegations. It asks the respondent to classify each fact: agreed, denied, or put to proof. It then asks targeted questions, probing the points that matter to the legal analysis.
From that material, Leonidas drafts the response. Numbered paragraphs. Restrained pleading. Admissions where appropriate. Denials with brief reasons. The claimant put to strict proof where the respondent cannot admit or deny. The output is a Word document, marked DRAFT, ready for solicitor review.
Leonidas is in prototype form. The AI is live. The interview uses the same reasoning that the production platform will use. The drafted response is produced from the respondent's answers in real time. Allow 15 to 20 minutes to complete the process.
Anita is in development. When complete, it will speak to witnesses early in the litigation and create litigation-privileged proofs of evidence, using the witnesses' own words. No more waiting until years after events to find out what happened.
In conventional litigation, witness statements are prepared months or years after the events in question. A solicitor telephones the witness, often years after the relevant events, and asks them to recall what happened. The witness remembers imperfectly. The solicitor writes it up in language the witness would never use. The statement is exchanged shortly before the hearing. The judge reads it and wonders why every witness for the defendant/respondent sounds the same.
The problem is not just quality. It is timing. The defendant/respondent's legal team often does not know what its own witnesses will say until weeks before the hearing. Tactical decisions about settlement, disclosure, and case strategy are made without a clear picture of the evidence. Cases are lost because the defendant/respondent's account was never properly recorded at the outset.
PD32, paragraph 18.1 requires that a witness statement must, if practicable, be in the intended witness's own words. Practice Direction 57AC, which governs trial witness statements in the Business and Property Courts, reinforces this at paragraph 3.3 and requires the witness to confirm that the statement sets out their personal knowledge and recollection in their own words.
In practice, this requirement is routinely breached. Solicitors draft statements in their own language. The witness's voice is lost. The statement reads as if written by a lawyer, because it was. Judges notice. Credibility suffers.
Anita will address this directly. Because it will record what the witness actually says, in the way the witness actually says it, the resulting proof will comply with PD32 paragraph 18.1 by design. The witness's own words will be preserved from the outset, not reconstructed months later by a solicitor working from incomplete notes.
Anita will conduct structured interviews with each witness as soon as the claim is received. The interview will follow protocols designed by barristers who have spent years proofing witnesses for tribunal hearings. It will identify the documents the witness can speak to, the decisions they made or observed, and the conversations they had.
The witness will choose how to give their account. Anita will work across three channels: a text-based interview on screen, a phone call using ElevenLabs-enabled voice, or a Teams call with an AI agent. The witness picks whichever suits them. The output will be the same: a structured proof in the witness's own words.
Anita is in the design phase. The witness proofing protocols are being written by the founders, drawing on their experience of proofing witnesses for employment tribunal hearings and High Court trials. Anita will integrate with Leonidas, using the fact classifications and case analysis already gathered during the defence drafting process to structure the witness interviews.
A grievance comes in. HR sends the complainant to Jesmond. What follows is a structured, fair, and thorough process that produces better outcomes in less time, with a complete documentary record from start to finish.
A formal grievance triggers a process that most HR teams dread. The complaint arrives as a long, unfocused letter. HR must work out what the actual issues are. A grievance hearing is arranged. Witness interviews follow. At the end of it, someone has to write up the outcome, often weeks later, from memory and sparse notes. The process is slow, inconsistent, and produces documentation that would not survive scrutiny at tribunal.
The cost is not just time. Badly handled grievances cause legal risk. Jesmond fixes this from the inside out.
For organisations handling dozens of grievances a year, Jesmond provides a dashboard view of every active grievance across the business. HR sees the status of each case, the issues raised, the witnesses identified, and the timeline. Grievance managers see their cases and the materials Jesmond has prepared for them.
The platform captures institutional data on grievance patterns: which issues recur, which departments generate complaints, and where the process breaks down. That data is not available from a filing cabinet full of outcome letters written by different managers in different styles.
Jesmond is in the design phase. The grievance protocols are being written by the founders, drawing on their experience of advising respondents on grievance processes and the tribunal claims that arise from them. Jesmond will integrate with Leonidas: where a grievance leads to a tribunal claim, the entire grievance record, including the transcripts, the evidence, and the outcome, feeds directly into the defence drafting process.
The DefendThis platform sits inside the organisation defending the claim. By the time external lawyers are instructed, the substantive work has already been done. The five steps below run in sequence.
The premise of DefendThis is that the slowest part of defending any claim is not the drafting. It is the gathering of facts from the people who hold them. Witnesses are dispersed, busy, and often reluctant. By the time their accounts have been collected, weeks have passed and the response deadline is close.
By embedding fact-gathering into a guided platform, that lag is removed. The witness account, the documents, and the legal analysis are produced in parallel rather than in series.
The output is a working first draft. Internal lawyers can finalise it in hours. External firms, when instructed, start from a complete file rather than a blank page.