Newcode vs Wordsmith: how they compare in 2026
Newcode and Wordsmith both sell AI workspaces to legal teams. Newcode is a layer across the systems a firm already runs, while Wordsmith is a front door that triages and resolves legal requests from the business. The gap between them is what each publishes about client data. Wordsmith names OpenAI, Anthropic and Google with the contract under which each retains nothing, and states that customer data trains no model. It addresses privilege and ABA Formal Opinion 512 directly and commits to breach notice within 72 hours. Newcode's security detail sits in a Vanta trust portal. Newcode's counterweight is reach and control. It connects to iManage, Outlook, SharePoint and hundreds of other systems through Model Context Protocol, retrieves within walls the firm defines, and offers local deployment as well as cloud. It names Kirkland & Ellis, DLA Piper and Reed Smith among its customers.
Which one fits
Choose Newcode if
- Your firm wants AI across the systems it already runs rather than another repository. Newcode connects to iManage, Outlook, SharePoint, OneDrive, Teams and the Office applications, describes reach into more than 750 Model Context Protocol servers, and holds no documents of its own.
- You want to build your own agentic workflows. Newcode's Aurora is a visual builder in which a firm assembles workflows from agents, prompt templates and a model selector, then deploys them, with retrieval bounded by the taxonomy, labels and walls the firm defines.
- You need a local deployment option. Newcode offers cloud or local deployment with matter level configuration and granular permissions, and names Kirkland & Ellis, DLA Piper, Reed Smith, The Bar of Ireland and Norwegian public agencies among its customers.
Choose Wordsmith if
- Your business floods legal with requests through Slack and email. Wordsmith triages each request by priority, jurisdiction and commercial exposure, answers routine ones, and routes the rest to a named lawyer with context from connected systems.
- You must tell clients or the board which models see your data. Wordsmith names OpenAI, Anthropic and Google, each through a paid enterprise API, and the specific contract under which each applies zero data retention, and hosts on AWS with EU residency by default.
- You want answers you can verify cell by cell. Wordsmith's research answers link to the underlying legislation across more than 130 jurisdictions, and its Reports link every extracted value to an exact clause and page for a reviewer to approve or flag.
At a glance
All 15 axes, side by side
The same grid applied to every vendor in the index, graded from public sources. Hover a grade to see what the letter means on that axis.
AI Centrality
How much of the product is actually AI. Whether the machine learning is the mechanism the buyer is paying for or a feature layered onto conventional software, and whether the vendor is specific about which is which.
Newcode holds no documents of its own. It connects to the systems a firm already runs, and everything it adds on top is model driven. Nova retrieves, analyzes and drafts with referenced answers. Aurora is described as the AI native intelligence engine at the heart of the platform, a builder in which a firm assembles agentic workflows from agent nodes, prompt templates and a language model selector. Newcode describes retrieval as an agent that plans, searches, evaluates and iterates until the context is right, rather than a lookup. The published capability list names workflow orchestration, contract review, case law analysis, knowledge ingestion, data structuring, document analysis, structured outputs, legal reasoning and governed execution. Newcode calls the product an AI harness rather than a workspace with AI in it.
The product runs a four step loop of receive, route, resolve and record, and models carry the middle two. Triage classifies an incoming request by priority, jurisdiction and commercial exposure and sends it to the right lawyer, and agents answer or draft without one where the work is routine. Agents respond in Teams, Slack or email and also run on a schedule for matters that repeat. Review checks inbound paper against the playbook with more than 100 checks, explains why a clause breaks policy using past negotiated positions, and on a returned draft reviews only what changed. Repositories scan each contract as it arrives for risks such as unlimited liability, and extract renewals, notice periods and obligations with alerts 90 days before expiry. Research gives jurisdiction aware answers that adapt to each entity's risk profile through Company Profiles, and Personal Memory folds a user's edits back into later answers.
Citation Accuracy and Hallucination Disclosure
Whether the vendor publishes measured accuracy on citations and assertions, grounds output to primary sources, and says plainly what its system does when it does not know. Legal has a documented public record of fabricated citations reaching filed briefs, so an untested claim of accuracy is not evidence.
Nova returns answers with references, grounded in the firm's own data or in verified outside sources, Newcode says. The retrieval agent plans, searches, evaluates and repeats until it has the right context, guided by the firm's taxonomy, labels and walls. The sources are the firm's own documents and outside sources, and answers carry references to what they rest on. No accuracy figure, evaluation, test set, error rate or failure mode is published, and there is no accuracy or benchmark page. Hallucination is not addressed, and no customer agreement is published.
Research draws on primary legislation, databases and guidance from curated government, regulatory and authoritative sources across more than 130 jurisdictions in 60 countries. The sources update continuously, and live web results fill in recent developments. Companies House, EDGAR and Open Corporates connect for live checks on entities and officers. The assistant returns cited summaries linked to the underlying legislation. Reports links every extracted value to its exact clause, page number and surrounding context, and a reviewer clicks a cell to check it, then approves and locks it or flags it. How a conflict between a curated source and a web result is settled is not described, and no measured accuracy is given.
Autonomy and Oversight Model
What the system decides on its own, what a lawyer must approve, and whether the vendor documents where the review point sits. A tool that drafts under review and a tool that files without one are different products and different risks.
Controls on offer include configuration by matter, granular permissions, audit, firm defined taxonomy, labels and walls, the ability to choose the model and set guardrails without waiting for a release, and governed execution. The interface shows an expandable thought process against each output. Aurora is sold on deploying agent workflows at scale to automate complex processes, and the interface shows an Auto toggle on a running task. Newcode does not say when the system acts without a person. Nothing says which outputs need review, what happens when an agent is wrong, whether generated work is marked, or what guardrails can be set to. No written commitment says the models work alongside a supervising lawyer.
Work reaches a lawyer only when it needs judgment. Intake rules, written in plain language, decide which requests become tasks. Each task is routed to an agent, a lawyer or a team by expertise, priority, region or business unit, and can be assigned, tracked and moved to resolution. Each named agent carries its own limits. The NDA agent returns routine NDAs and sends non standard positions to legal for sign off. The procurement agent handles routine purchases and escalates any above a value threshold. The privacy agent sends anything outside its playbook to legal. The customer chooses which low risk work runs end to end with no person involved, and other matters escalate with their context and reasoning. Agent actions are logged with a timestamp and source. Review surfaces are concrete. In Reports a reviewer checks each cell against the cited clause, then approves and locks it or flags it. In Word and Google Docs each proposed redline can be accepted, rejected or changed. What follows a wrong answer is not described.
Operational and Outcome Evidence
Named, dated evidence that the product works in production at real firms or legal departments. Case studies with figures and identified customers count. Unattributed testimonials and launch announcements do not.
Customer names run across Newcode's home page in text and logos: Kirkland and Ellis, DLA Piper, Reed Smith, Dickinson Wright, Wolters Kluwer, Robinson and Cole, The Bar of Ireland, Wiersholm, Haavind, Awilhelmsen, Ræder Bing, Simonsen Vogt Wiig, Ekko Advokatfirma and Helton. It also names Norwegian public bodies including Helsetilsynet, Meteorologisk institutt and the Statsforvalteren offices, and claims more than one hundred law firms and enterprises. In one named testimonial, Preben Brecke, Managing Partner at Haavind, says the product helps structure transactions, sharpen analysis and reach better decisions in private equity, M&A and complex disputes. No figure, date or described outcome accompanies any customer, and no case study is published.
Several named customers give figures. Trustpilot's story puts the cut in contract review time at 85%, with its chief trust officer, Anoop Joshi, quoted. Eye Security's general counsel, Stacey McKenzie-Vass, reports risk scans of historical customer agreements taking hours instead of a week, about 90% less time. At Cint, senior legal operations specialist Isa Romera describes 40% of a lawyer's week returned and counterparty paper reviewed the same day across four jurisdictions. Belron's account of April 2026, from senior corporate counsel Christopher Allen and legal operations vice president Jessica Vander Ploeg, counts 6,843 interactions in March. Its average of 90 minutes saved a day comes from a quarterly user survey, and its 60% to 80% efficiency gain is the average users reported during the trial. The customers page carries 17 stories, among them Canva, Trip.com, Bynder, the Financial Times, Multiverse, Typeform, Fugro, DataSnipper, Contentsquare, CoachHub and Preply. Kantar's Trevor Brannan describes measuring value with Usage & Value, and Ebury chose Wordsmith in November 2026 for a single intake route, starting with Jira. More than 500 companies use the product by Wordsmith's count of June 2026. The figures are the customers' own accounts or survey results.
Privilege and Confidentiality Posture
How client confidences are handled: attorney client privilege and work product treatment, segregation of one client matter from another, whether client data trains any model, and what the vendor commits to in writing rather than in marketing.
Retrieval follows firm defined taxonomy, labels and walls, according to Newcode's home page, and a firm gets governed, auditable configuration by matter with granular permissions over what each team can reach. No customer agreement, master services agreement or data processing addendum is published. The site footer links only a cookie policy and a privacy policy. The privacy policy says in its first line that it covers personal data collected through the website or contact form, so it does not reach matter data. Nothing is published on training, privilege, work product, or what any model provider keeps. The site's Security page links to a trust center on Vanta.
Clause 3.2.1 of the terms of 31 October 2024 bars Wordsmith from training any large language model or other machine learning model on customer inputs. The security page adds that customer data is used for no purpose beyond service delivery. OpenAI, Anthropic and Google are each named with the contract under which zero data retention applies, and prompts and outputs are not stored, logged or used for training by any provider. The security page says the product is designed to maintain attorney client privilege and that its architecture aligns with ABA Formal Opinion 512. Clause 11 binds each party to keep the other's confidential information. Data is kept separate from other customers, though how that separation works is not described.
UPL and Professional Responsibility Posture
Whether the vendor is clear that it supplies a tool rather than legal advice, who its audience is, and how it addresses unauthorized practice of law, competence and supervision duties, and jurisdiction limits. ABA Formal Opinion 512 is the reference point. Where the advice line is not the duty a product raises, the axis is read through the nearest professional duty it does raise: judicial conduct rules and the reviewing duty for products sold only to courts, and the duty to bill for time actually spent for products that draft time entries.
No published material addresses the advice line. There is no statement that Newcode is not a law firm and no disclaimer that output is not legal advice. Nothing covers the professional judgment needed before generated work is used, any jurisdiction limit, or supervision and competence. No terms of service are published, and the privacy policy covers only the website. The published use cases are legal work itself. They include drafting engagement letters, extracting court dates and deadlines from a scheduling order, conducting discovery, preparing trial submissions, giving strategic advice, and handling government applications, complaints and administrative proceedings.
Clause 9.3 of the terms provides output for general information only, and clause 9.7 puts decisions made in reliance on the service, including legal advice, at the customer's own risk. The security page says the product supports professional judgment rather than replacing it and points teams to ABA Formal Opinion 512 and local bar guidance on disclosing AI use, noting that obligations vary by jurisdiction. The product is built to resolve routine requests from sales, procurement, HR and security without a lawyer, under rules the legal team writes. The lawyer's own competence and supervision duties are not addressed, and no jurisdiction limit is set on what agents answer.
AI Governance and Bias Disclosure
Published governance over model behavior: who owns it inside the vendor, what is tested before release, and what is disclosed about disparate output across matter types, parties, or populations.
No governance position is published. There is no responsible AI page, principles statement, named owner for model behavior, testing before release, management system or ISO 42001 certification. Governed execution, guardrails and configuration by matter are controls a customer sets over its own deployment, not disclosure of how Newcode governs its models. The privacy policy's clause on automated decisions covers only website data. Nothing says who is accountable for model behavior when the firm has picked the model and built the workflow.
No governance position for model behavior is set out. Nothing names a person or role accountable for what the models produce, describes testing before release, states responsible AI principles or reports uneven output across matter types or counterparties. The security material covers information security only. Usage & Value reports what work runs through the product and its estimated value, which measures use rather than quality.
AI Safety and Data Stewardship
Retention, deletion, access control, and what happens to prompts and documents after they are processed. Whether the vendor states its subprocessors and its incident practice, or leaves the buyer to assume.
Newcode states granular permissions, configuration by matter, firm defined labels and walls that limit what retrieval can reach, audit as a built in feature, and a choice of cloud or local deployment. No retention, deletion, incident practice or subprocessor list is published for customer data. The privacy policy gives retention periods of twelve months for contact form messages and ninety days for technical logs, sets out GDPR legal bases, names transfer mechanisms and lists three website providers (Vercel, Resend and Cloudflare Turnstile). Its first line limits it to data collected through the website. No customer agreement is published to cover the rest.
Data is encrypted with AES 256 at rest and TLS 1.2 or higher in transit, and internal access follows least privilege. Single sign on runs through Okta, Azure Active Directory and Google Workspace over OIDC and SAML 2.0, with multifactor authentication and role based access control. An outside firm tests the platform every year, and under a formal incident response plan a confirmed breach is reported to affected customers within 72 hours with details and remediation steps. Customers control retention. Data can be deleted permanently at any time, all customer data is deleted when the contract ends, and custom retention policies can match internal governance. The subprocessor list, provider policies, data processing agreement and incident response plan sit in the Vanta trust portal, which is closed to automated reading.
AI Liability and Recourse
What the vendor stands behind contractually when its output is wrong. Indemnities, caps, carve outs, insurance, and whether any of it is published or only reachable through a negotiated agreement.
No published document addresses who bears the loss when the system is wrong. The site footer has two legal documents, a cookie policy and a privacy policy, both on a single legal page. No terms of service, master services agreement, customer agreement or data processing addendum is published, so no indemnity, liability cap, warranty, disclaimer, service level or insurance statement is public before a sales process. The privacy policy covers website visitors and allocates no liability.
Clause 9.1.2 of the terms caps Wordsmith's total liability at the higher of £10,000 and the fees paid in the 12 months before the claim. Clause 9.1.1 excludes listed categories of loss, and clause 9.2 keeps liability for death or personal injury from negligence, fraud and anything the law does not allow to be excluded. The indemnity in clause 8 runs from the customer to Wordsmith, covering use of the service and breach of the terms. Clauses 9.3, 9.4 and 9.7 disclaim the accuracy of output and leave decisions made in reliance on it to the customer. No indemnity to the customer, warranty on output or insurance position is offered. English law and the courts of England and Wales govern.
Practice Systems Integration Depth
How deeply the product reaches into the systems legal work already lives in: document management such as iManage and NetDocuments, Word and Outlook, contract lifecycle management, matter management, e-billing, and court filing systems.
Newcode connects to the systems a firm already uses rather than replacing them. The named systems are iManage for document management, Outlook, SharePoint, OneDrive, Teams, and Word, Excel and PowerPoint. Newcode says it reaches these through access to more than 750 Model Context Protocol servers. A published example drafts an engagement letter from a matter in Outlook, using past letters in SharePoint as fee benchmarks. The assistant includes a document management search tool, and review, redlining and feedback happen directly in Word. Newcode says context follows the matter rather than the tool. No field mapping, sync direction, trigger or permission model is documented, no developer or configuration documentation is published, and the 750 figure is a count, not a list.
Microsoft Word carries playbook reviews, redlines, precedent, translation and consistency checks inside the document. A Google Docs add on, launched on 28 October 2026, reviews against playbooks and applies redlines as tracked changes. Slack and Microsoft Teams host intake and agents, and Teams admins link agents to channels and set which ones answer direct messages. In Outlook and Gmail the product summarizes threads, drafts replies from the playbook and starts reviews from attachments. Wordsmith has also been in ChatGPT's app directory since 29 October 2026. The Connector Library, launched on 3 November 2026, groups connectors by job. Document stores include iManage Work, SharePoint, Box, Egnyte, Google Workspace, Datasite and Dropbox. Contract and signing tools include Ironclad, SAP Ariba, DocuSign CLM, DocuSign, Adobe Acrobat Sign, Juro and PandaDoc. Salesforce, Dynamics 365, Workday, Snowflake, HubSpot and Attio supply business context. ServiceNow, Zendesk, Freshservice and Slack carry requests, and Diligent Entities, Carta, LawVu, Moonlit and PERSUIT serve specialist legal work. Some connectors are read only, and others act. A redline can be saved back to Box, Egnyte or SAP Ariba as a new version. A ServiceNow or Zendesk ticket can get a reply with a document attached, a draft can start an Ironclad workflow, and SharePoint contract registers can be updated as work moves. Access follows the connector settings and approval rules the customer sets. Custom connectors, MCP and an API reach other systems, and Zapier and n8n open more than 8,000 more. Setup steps and the full read and write split for each connector are not described on the open site.
Deployment Model and Data Residency
Where the software runs and where the data sits. Multi tenant cloud, single tenant, private deployment, on premises, and whether region of residence is a published option or an enterprise conversation.
Deployment can be cloud or local, a self hosted route alongside hosted cloud, with configuration by matter and granular permissions over what each team can reach. No region is named for the hosted option, no cloud provider is identified, and no residency commitment is published. The only places named are company offices in Oslo, Stockholm, Dublin, New York and Palo Alto. The privacy policy's transfer section covers only website personal data.
Hosting is on AWS, with EU data residency by default and the United States available, on infrastructure certified to ISO 27001, SOC 1, 2 and 3 and PCI DSS Level 1. Data stays in the customer's chosen region unless the customer consents to a move. Each customer's data is kept separate from every other. No single tenant, private or on premises option is offered. Prompts go to OpenAI, Anthropic and Google, and where that processing runs relative to the chosen region is not stated.
Security Certifications and Trust Center
Independent attestation a buyer can pull without a sales call: SOC 2, ISO 27001, penetration test summaries, a trust center with current reports and named scope rather than a badge image.
The site footer links Security to a trust subdomain hosted on Vanta. The site's own pages name no standard, certifying body, examination period, scope or report, and claim no certification. No certification badge appears on any page.
SOC 2 Type II is stated, with an independent audit each year, alongside GDPR compliance and annual independent penetration testing. The AWS infrastructure underneath is certified to ISO 27001, 27017 and 27018, and no ISO certification is stated for Wordsmith itself. Full documentation sits in a Vanta trust portal linked from every page, which is closed to automated reading. No auditor, report period or scope appears on the open site.
Model Supply Chain Disclosure
Which models sit underneath, whose they are, where they run, and whether the vendor commits to telling customers when that changes. A legal buyer inherits every dependency it cannot see.
Buyers are offered the language model that works best for a firm, told they can choose models and set guardrails without waiting for a release, and shown a model selector inside an agent node in Newcode's workflow builder. No provider, model family, version or hosting arrangement is named, and no commitment to notify customers when the available models change is published. One interface illustration shows a model name, and nothing else confirms it.
Three model providers are named, each with the contract that sets its terms. OpenAI runs under its API Data Usage Policy, Anthropic under its Commercial Terms and Google's Gemini under the paid API Gemini Terms, all through paid enterprise APIs with zero data retention. Research also pulls live web results. No specific model or version is named, and no commitment covers telling customers when a provider or model changes.
Commercial Transparency
Whether a buyer can learn what this costs without entering a sales process: published rates, the unit being charged, what sits behind an enterprise tier, and what implementation adds.
No pricing is published at any level, including the unit of charge. There is no pricing page in the navigation (solutions, academy, customers, security, about, news, careers and contact) or in the footer, which has only the cookie and privacy policies. Every commercial link leads to booking a demo. Nothing says whether the platform is licensed per user, per firm, per workflow, per agent run or by usage. No tiers, minimum commitment, term or implementation cost are published. No terms of service or master agreement is published, so no payment terms are public either.
No price is listed at any level. Two plans are sold, Pro and Enterprise, and buyers start with a 30 minute demo. A free trial includes a limited version of Reports, and full Reports needs a Pro or Enterprise plan. Enterprise accounts get white glove implementation, and every paying customer gets support. The terms set the rest of the shape. Fees are fixed in an order form and invoiced monthly. A subscription runs for an initial term and renews automatically, and either party can end it at the term's end with 30 days notice. Wordsmith gives 30 days notice of a fee change. A material change to the service lets the customer cancel with a prorated refund. An ROI calculator estimates savings, and Usage & Value prices the work done at the customer's own in house and outside counsel rates.
Firm and Practice Coverage
Who the product is actually built for. AmLaw, midlaw, small firm and solo, in house departments, government and courts, and which practice areas are supported rather than merely claimed.
The pitch splits across five buyer segments. Law firms are addressed on billable capacity and firm wide standards, and in house teams on standardizing routine work, traceability and turnaround. Insurers are addressed on matching claim facts to policy wordings and endorsements and surfacing coverage grants, exclusions and conditions precedent. Banks are addressed on analyzing facility, loan and master agreements at volume, with review and redlining in Word, and government on handling applications, complaints, investigations and administrative proceedings. The customer list runs from international firms to a bar association to national public agencies. Named capabilities include fund formation, case law analysis and jurisdictional comparison. Newcode states no jurisdiction limit, despite a Nordic base and US offices, and no firm size range, practice area or matter type the product does not suit. Supported languages are not listed.
Only in house legal teams are served, with workflows designed around general counsel, legal operations and in house lawyers rather than law firm billing. Solution pages cover general counsel and legal operations, the procurement, sales, security and HR teams that send requests, and company sizes from startups to enterprises. Legal coverage is stated by jurisdiction, more than 130 across 60 countries, and the site runs locale versions in several languages. Practice coverage is expressed through the requesting team and contract type rather than as areas of law, and government and court use are not addressed.
The 12 legal signals, side by side
Recorded rather than graded. These are the questions a practitioner has to answer before a tool touches a client matter, and the answers are taken from public material only.
Client Data in Training
Can material a lawyer puts into this product be used to train a model?
No customer agreement is published, and no page states a position on training. The site footer links a cookie policy and a privacy policy on one legal page, and no terms of service, master services agreement or data processing addendum. The privacy policy says in its first line that it covers personal data collected when a visitor uses the website or contact form, so it does not reach matter data and says nothing about model training. No product page says whether customer content trains models, and no opt out or setting is described.
Clause 3.2.1 of the terms of 31 October 2024 commits Wordsmith not to train any large language model or other machine learning model on customer inputs. The security page adds that customer data is not used to improve the platform for other customers. OpenAI, Anthropic and Google process prompts under contractual zero data retention and do not train on them.
Prompt and Output Retention
How long does the product keep what a lawyer typed, and can that be set to zero?
Nothing published says how long anything the platform handles is kept. The privacy policy gives retention periods of normally twelve months for contact form submissions and related messages and ninety days for technical and security logs, longer where law requires. That policy covers only personal data collected through the website and contact form. Nothing addresses retention of documents retrieved from a firm's connected systems, the context gathered by the retrieval agent, prompts, or the outputs and workflow runs Aurora produces. No commitment to delete or return customer material is published, and no customer agreement is published to carry one.
Customers keep control of their data, can delete it permanently at any time and can set custom retention policies to match internal governance. All customer data is deleted when the contract ends. No default retention period is given for prompts or outputs, and the model providers keep nothing after processing.
Ethical Walls and Matter Segregation
Does retrieval respect the firm’s ethical walls, or can the model read across them?
Newcode's home page says the retrieval agent plans, searches, evaluates and repeats within firm defined taxonomy, labels and walls, and offers governed, auditable configuration by matter with granular permissions over what each team can reach. Walls apply to what the retrieval agent can see, not only to what a user can open. Nothing says whether the platform is shared or single tenant, how a wall is configured or audited, what happens when an agent moves through connected systems with their own permissions, or whether walls hold across the more than 750 outside connections the product claims. No administrator documentation is published.
Each customer's data is kept completely separate from every other customer's, with no mixing. Single sign on, multifactor authentication and role based access control sit on top, and playbooks and templates are shared across an organization only when a user chooses. How the separation between customers, or between teams inside one, is enforced is not described.
Third Party Request and Subpoena Notice
If someone subpoenas the vendor for a firm’s data, does the firm hear about it first?
No published term or policy addresses third party requests for customer data, and no master agreement is published. The privacy policy covers the subject for its own data, listing public authorities, courts and regulators among recipients where law requires, but it commits to no notice, reserves no discretion over notice and publishes no transparency report. It covers only website and contact form personal data, not a firm's matter content.
Nothing says what happens when material held or reachable through the platform is demanded by subpoena, court order or regulator.
Clause 11.4 of the terms lets a party disclose confidential information when a law, regulator or court requires it, and where the law permits it gives the other party as much notice as possible. The disclosing party also takes account of the other party's reasonable requests about what is disclosed. The duty is mutual and survives termination. No transparency report is offered.
Primary Law Corpus Provenance
Where does the law in this product come from, and does the vendor have the right to use it?
Newcode retrieves from the systems a firm already runs rather than from its own database, and calls the result firm owned intelligence, so the primary material behind an answer is the customer's. Nova also searches legal and outside sources and returns answers grounded in the firm's data or verified outside sources, and case law analysis and legal research are listed capabilities. No publisher, database, jurisdiction set or licensing position is stated for outside legal material, and Newcode does not say who verifies the outside sources or against what. A jurisdictional comparison tool is offered without naming the jurisdictions it covers.
Research covers more than 130 jurisdictions across 60 countries, with primary legislation, databases and guidance from hundreds of curated government, regulatory and authoritative sources. Those sources update continuously, and live web results fill in recent developments. No individual source of legislation or case law is named, and no license basis is given. Companies House, EDGAR and Open Corporates are named, as registries for checking entities and officers.
Good Law Verification
Does the product tell you when the authority it just cited has been overruled?
Case law analysis and legal research are both listed capabilities, Nova has a legal research tool and a jurisdictional comparison tool, and answers are said to be grounded in verified outside sources. No citator, treatment signal, currency check or good law check is described, and nothing says what verified means or who does the verifying.
Answers come back cited and linked to primary legislation, from sources that update continuously, with live web results for recent developments. Whether a cited provision or decision still stands is not flagged, and no citator or treatment signal is part of the product.
Refusal and Uncertainty Behavior
What does the product do when the answer is not in the corpus?
No published material describes what the system does when it cannot ground an output. The retrieval agent is described as planning, searching, evaluating and iterating until the context is right. Nothing states what happens when the right context is never found, whether the agent stops, reports the gap or answers anyway, and no confidence indicator, abstention path or no answer state is described. A reasoning trace is visible in the interface as an expandable thought process, which shows what the system did rather than how certain it was.
Nothing addresses behavior when a connected system is unreachable or a firm's walls exclude the material an answer would need.
Triage sends a request to a lawyer when it needs judgment, and intake rules can hold a task for approval before it starts. In Reports a reviewer can flag any extracted value for review. No behavior is described for a question the system cannot answer from its sources, and no confidence measure is shown.
Fabricated Citation Record
Does a public court record exist addressing fabricated or hallucinated legal citations in output from this product?
The AI Hallucination Cases database maintained by Damien Charlotin records no case under the product names Newcode, Nova or Aurora or the corporate name Newcode.ai AS. No court order, opinion or disciplinary record naming the product or the company is on the public record. Nova and Aurora are common product names across the software industry.
The AI Hallucination Cases database maintained by Damien Charlotin, which records court decisions worldwide that address hallucinated AI content and the tool involved where known, has no entry naming Wordsmith or Wordsmith AI Ltd. Its products serve in house legal teams and the business functions they support.
Bar Guidance Alignment
Has the vendor engaged in public with the ethics opinions its buyers are bound by?
No published material engages with bar or ethics guidance at any level. No bar association, law society, conduct rule, ethics opinion or regulator guidance is named or referred to, and nothing explains what a firm must do to meet its own supervision and competence duties when agent workflows it built are running across its documents. There is no advice disclaimer or statement that the company is not a law firm either. The Bar of Ireland is a named customer.
The security page cites ABA Formal Opinion 512 twice, saying the product's architecture aligns with it and, in a FAQ on disclosing AI use, telling teams to review it with local bar guidance. That answer notes that some bodies recommend disclosure for significant AI use while others do not require it for internal tools. No opinion from another jurisdiction is named.
Billing and Fee Posture
Does the vendor address what happens to the bill when the work takes an hour instead of six?
The law firm proposition is published as reducing repetitive administrative work and expanding firm wide billable capacity, alongside compressing turnaround times, protecting quality at volume and expanding capacity so teams can deliver more. No record per matter of AI assisted work is described as available, no guidance on fee or disclosure treatment is published, and nothing addresses what a client is told when research, drafting or due diligence on their matter was produced by an agentic workflow the firm configured. Matters are described as traceable and execution as auditable, offered for governance rather than for disclosure.
Wordsmith is sold to in house legal teams, with solutions for general counsel and legal operations and for the procurement, sales, security and HR teams they support, from small businesses to enterprises. Named customers include Canva, Trip.com, Bynder, Belron, Trustpilot and Selfridges. Its savings claims are aimed at those teams. A customer story says Trustpilot cut contract review time by 85%, and the home page counts hours saved per lawyer each week and outside counsel spend. Wordsmith Usage & Value puts a number on what the product delivers to the legal department.
Outside Counsel Guideline Readiness
Can a firm get this vendor through a client’s AI clause without a bespoke negotiation?
No subprocessor list for the platform is published, no model or model provider is named, no cloud provider is identified for the hosted option, and there is no data processing addendum, consent pack or notice material. The only subprocessors published are three website providers in the privacy policy (Vercel, Resend and Cloudflare Turnstile), which serve the marketing site and do not touch client content. The platform is sold on model choice and on reaching more than 750 outside systems, and none of the parties that may see client content through them is disclosed.
The three providers that see customer content, OpenAI, Anthropic and Google, are named on the security page with the contract that sets each one's zero data retention terms, readable before any agreement. The full subprocessor list, data processing agreement and incident response plan sit in the Vanta trust portal. No client facing disclosure template is offered.
Court Disclosure Support
If a judge’s standing order requires an AI disclosure, can the product produce one?
Newcode says the platform keeps every matter traceable, offers governed and auditable configuration by matter, and shows a thought process against outputs, so a firm can see what a workflow did and under what settings. Nova's answers carry references to the sources retrieved. No model or version is shown against any output, although the model can be chosen and may differ between workflows and over time, so the record does not show which system produced what.
Nothing marks generated work once it leaves the platform as a document, no human verification is recorded, and no export is described for producing any of this to a court, client or regulator. No disclosure template or guidance is published.
Reports links every extracted value to its exact clause and page, and records whether a reviewer approved and locked it or flagged it. In Word, each proposed redline is recorded as accepted, rejected or improved. The model behind a passage is not shown, and no export for a court filing is offered.
The questions both sides leave open
Derived from the records above rather than written, so it cannot favor either vendor. Take these into both conversations and ask each side the same question.
- AI Governance and Bias Disclosure
- Good Law Verification
- Refusal and Uncertainty Behavior
Product updates
The latest logged changes from each vendor, newest first, each linked to its source.
Newcode
-
Product / capability
Newcode launched Configurable Skills, which turn a firm's processes, playbooks and risk positions into reusable agent capabilities written in plain language.
Source
Wordsmith
-
Integration / interoperability
Wordsmith released an add on for Google Docs, listed on the Google Workspace Marketplace, that brings its contract tools into the editor.
Source -
Integration / interoperability
Wordsmith joined the iManage Technology Partner Program.
Source -
Safety / governance
Wordsmith shipped enterprise controls aimed at extending AI assisted legal workflows beyond the legal team to other business units.
Source
In summary
Newcode
Newcode is an AI workspace from Newcode.ai AS of Oslo, with offices in Stockholm, Dublin, New York and Palo Alto. It connects to the systems a legal team already runs, including iManage, Outlook and SharePoint, rather than holding documents itself. Its assistant, Nova, searches the firm's material and external sources and returns referenced answers, and Aurora lets a firm build and deploy its own agentic workflows with a choice of model. According to the AI Legal Index, it offers cloud or local deployment and names Kirkland & Ellis and DLA Piper among customers. Every commercial link leads to booking a demo.
Wordsmith
Wordsmith, founded in 2023 in the United Kingdom, is a legal front door for in house teams. It captures requests from Slack and email, triages them by priority, jurisdiction and commercial exposure, and resolves routine ones or routes them to a lawyer. It also offers playbook contract review, drafting, agents and research across more than 130 jurisdictions. According to the AI Legal Index, it names OpenAI, Anthropic and Google with zero retention terms, states SOC 2 Type II and EU residency by default, and names Trustpilot, the Financial Times and Belron among customers. No price is listed, and its Pro and Enterprise plans are quoted after a demo.
Questions buyers ask
Newcode vs Wordsmith: which is better for an in house legal team?
Wordsmith names the model providers behind it and sets out how it handles client data and its professional responsibility position. Newcode connects across a firm's existing systems and can run as a local deployment. Teams that must explain their AI supply chain to clients will find more to point to with Wordsmith. Teams that want the tool inside their own environment should start with Newcode. From the AI Legal Index, based on each vendor's own published materials as of September 27, 2026. No vendor pays for placement.
Which AI models does Wordsmith use?
Wordsmith's security page names OpenAI, Anthropic and Google, each used through a paid enterprise API, and names the contract under which zero data retention applies to each: OpenAI's API data usage policy, Anthropic's commercial terms and the paid API Gemini terms. Newcode lets a firm choose its model. From the AI Legal Index, based on each vendor's own published materials as of September 27, 2026. No vendor pays for placement.
Does Newcode work with iManage?
Yes. Newcode names iManage among the systems it connects to, alongside Outlook, SharePoint, OneDrive, Teams and the Office applications, and carries a document management search tool inside its assistant. Wordsmith joined the iManage Technology Partner Program on 2 September 2026. It reads content held in iManage through iManage's APIs, and it describes filing finished work back into iManage as planned for future releases. From the AI Legal Index, based on each vendor's own published materials as of September 27, 2026. No vendor pays for placement.
Does Wordsmith train AI on customer data?
Wordsmith's security page states that customer data is not used to train its models or to improve the platform for other customers, and a published article says customer data never enters any underlying model's training. That commitment is a policy statement on its public pages. From the AI Legal Index, based on each vendor's own published materials as of September 27, 2026. No vendor pays for placement.
Neither vendor paid for inclusion, placement or a grade, and neither reviewed this page before it published. Everything on it comes from public material on the dates shown. How the index grades.
Three points to weigh. Wordsmith's terms require the customer to indemnify it and cap Wordsmith's liability at the greater of £10,000 and the fees paid in the 12 months before a claim. Its terms also bar training any model on customer inputs, in clause 3.2.1. Newcode publishes no customer agreement and keeps its security detail in a Vanta trust portal, so its gaps reflect what is public, not a finding that its controls are weak. Neither vendor reviewed this page.