F
fynk
fynk is a contract management platform from fynk GmbH in Vienna, founded in April 2022 by a team that had built and sold the recruiting software Prescreen. It covers drafting from templates, negotiation, approval workflows, electronic signatures at simple, advanced and qualified levels with no envelope fees, and a searchable contract repository, for legal, finance, sales, procurement, HR, IT and operations teams. Its AI features include an assistant for questions across contracts, extraction of key terms and dates, review against a company's own rules with each clause marked green, orange or red and explained, drafting from templates, summaries and import.
An MCP server lets assistants such as Claude and ChatGPT draft from templates, file documents and start signature runs within a user's permissions. Prices are published per user: a free plan for one user, Essential at 19 euros and Business at 39 euros a user a month, each with monthly AI credits, and a custom Pro plan. The platform runs on AWS in the European Economic Area, with the AI service hosted at Hetzner in Germany, and fynk holds ISO 27001 certification issued by DQS. Named customers include Conrad Electronic, 1. FC Nürnberg and Unite.
Capability grades
All 15 axes, graded from public sources on the date shown. 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.
The models drive the capabilities fynk leads with, on a contract platform that would function without them. The platform underneath is templates, negotiation, approval workflows, electronic signatures at three eIDAS levels, a repository with reminders, and integrations into CRM systems, none of which needs a model. The features page lists seven AI functions: an assistant for questions across contracts, AI analysis that extracts key terms, AI review against a company's own rules, AI drafting, AI import, AI search and AI summaries.
AI review is described in three steps: set the rules, run the review, and fix the clauses it marks before the contract goes to legal review. The MCP server lets outside assistants draft from templates, file documents and start signature runs. Every plan, including the free one, includes the AI features, with monthly AI credits of 50, 150 or 250 depending on the plan and a custom allowance on Pro. The AI is part of the base product, with its use metered. Remove the models and fynk is still a working contract system, without the review and extraction it is now sold on.
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.
Review findings link to the clause they rest on and come with reasons, and no accuracy figure is published. The AI review page says each flag includes references that link directly to the relevant contract sections, reachable with one click, and that orange and red flags carry a written explanation against rules the customer sets. That is grounding a user can open, in the contract under review. Section IV of the terms for paid plans (version 2.3, 27 August 2026) states that the accuracy, completeness and suitability of AI results are not part of the contractual services and that the customer bears full responsibility for verifying them.
Section IV.8 adds that inaccuracies or gaps in results, including so called hallucinations, do not constitute a defect or a breach of contract. No benchmark, error rate or test of extraction or review accuracy is published on the features, AI review, legal team or security pages. The customer story figures, such as finding a contract in ten seconds rather than ten minutes, measure speed rather than accuracy.
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.
fynk states that its AI flags rather than decides and describes the review surface, without a control structure for the automated paths. The legal team page says AI in fynk does not decide: it highlights, surfaces risk, explains clauses and summarizes obligations, and the user stays in control of every decision. The AI review page describes a three color result for each rule, a written reason for orange and red flags, a link to the clause, and use before a contract goes to legal review.
The AI does not change the contract. Section IV of the terms puts the duty to verify AI results on the customer. Approval workflows with checkpoints, audit trails and version control are included in every plan. The automated paths have no stated limit. The MCP server lets outside assistants draft from templates, file documents and start signature runs, constrained by the user's permissions and logged in the audit trail.
AI import and analysis populate contract data without a stated review step. Nothing says what happens when an extracted date or term is wrong and a reminder or workflow runs on it.
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.
Named customers and named people are published with figures, without dates or methods. The customers page quotes Dr. Jörg Burkhardt of 1. FC Nürnberg saying that finding a contract used to take up to ten minutes and now takes ten seconds, and the legal team page quotes Elisa Prößdorf, legal solutions manager at Unite, on a manual task cut from one hour to two minutes. Further stories cover maven, where one person handles all contracting, fiskaly, Five 14, and Conrad Electronic replacing five tools with one, and the legal page also quotes legal and governance staff at Nioby and Falken Tyre Europe.
The home page reports 80 percent time saved, two minute contract processing and 33 percent faster contract creation, and says more than 3,000 teams use fynk, with no source, period or method behind the percentages. None of the stories carries a publication date or describes how its figure was measured. A buyer has named companies and named contacts to ask and a set of results it cannot assess on the page.
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.
The marketing says customer data is never used to train AI models, and the agreement reserves the right to train on it once anonymized. The security and GDPR page states that data is never used to train AI models and that the customer remains sole owner and controller of its data. Section VII.3 of the terms for paid plans (version 2.3, 27 August 2026) grants fynk a nonexclusive right, unlimited in time, to reproduce and process customer data to improve the user experience, provide feedback on AI functionalities, run statistical evaluations and publish anonymized industry statistics.
It says customer data may not be passed to third parties for these purposes or used to train AI models unless it has been anonymized first. The agreement governs, and it permits training on anonymized contract content. Section XI binds both parties to confidentiality for five years after the contract ends, with an exception for disclosure the law requires. Nothing published addresses privilege or work product, segregation between customers or teams, or what Azure AI Services and Mistral retain of what they process.
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.
A plain no advice position is in the terms, the service is limited to businesses, and the AI is framed as highlighting rather than deciding, without supervision guidance or jurisdiction limits. Section IV.8 of the terms for paid plans says the customer is aware that fynk's results and output do not constitute legal advice, and section II.3 limits the terms to business customers. The legal team page says the AI does not decide, highlights risk and explains clauses, and leaves every decision with the user.
The users, though, are not only lawyers: fynk publishes pages for finance, sales, procurement, HR, IT, operations and founders, and the maven story describes one person handling all contracting. Nothing tells those users when a red flag or an AI draft needs a lawyer, and the review checks a contract against rules the customer writes, so the quality of the result rests on rules that may be drafted by someone without legal training.
No page says which governing laws or contract types the AI review and drafting are built for. A company's legal team learns where fynk's responsibility ends and not how its own colleagues should use the tool.
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 AI governance position is published. Checked the home, features, AI review, legal team, pricing, integrations, about and security and GDPR pages, the terms for paid plans (version 2.3, 27 August 2026) and the subprocessor list on 2 October 2026, plus one search of the vendor's site for an AI policy or EU AI Act statement. The security page's AI section covers where AI processing happens, which providers are used and that data is not used for training, which are data protection subjects rather than governance.
The terms treat AI output as outside the contractual service and acknowledge hallucinations. Nothing published names who inside fynk is accountable for the AI's behavior or describes testing before a feature ships. Nothing sets out a risk classification under the EU AI Act for a company based in Vienna, or reports how review and extraction quality varies across languages, contract types or rule sets. The trust center at trust.fynk.com was not read on the date shown and may say more.
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.
Hosting, subprocessors, access, backups and deletion at the end of the contract are published; an incident practice and a retention period during the term are not. The security and GDPR page says data is stored on ISO 27001 certified AWS servers in the European Economic Area, transmitted over encrypted channels and backed up regularly to external servers. Only authorized and trained fynk employees have access to an account, and only for a specific event.
The subprocessor list, valid since 22 July 2026 with dated earlier versions archived, names ten processors with purpose and location. They include Hetzner in Germany for hosting the AI service, Microsoft Azure AI Services and Mistral in the EEA, and three with occasional processing in the United States. A data processing agreement is published. Section IX of the terms gives the customer three months after termination to export its contracts, after which the account is deactivated and its files and data deleted.
Not located: how long contracts, prompts and AI outputs are kept during the subscription, whether a customer can set that, and how and when fynk notifies a customer of a security incident.
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.
A limitation clause with a high cap is published, and it excludes exactly the exposure the AI creates. Section X.6 of the terms for paid plans (version 2.3, 27 August 2026, Austrian law, courts of Vienna) limits fynk's liability in each calendar year to three times the total remuneration the customer owes, and section X.4 excludes consequential damages. Section IV says the accuracy, completeness and suitability of AI results are expressly not part of the contractual services and that the customer bears full responsibility for verifying them.
Section IV.8 says inaccuracies and hallucinations in AI output do not constitute a defect or a breach of contract. A wrong extraction or a missed clause in an AI review is therefore not something the cap ever reaches. No indemnity runs from fynk to the customer; section X.11 runs the other way, requiring the customer to indemnify fynk against claims arising from infringements the customer is responsible for. Third party tool liability is excluded.
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.
Integrations into the CRM and automation systems a company team uses are named with what they do, short of implementation documentation. The integrations page names HubSpot, which syncs deal data to fill contracts with counterparty details and lets users create, track and sign contracts from HubSpot records through its workflow builder; Pipedrive, which fills contracts from Pipedrive accounts; Salesforce for contract and proposal generation and signature; Zapier, connecting thousands of other apps; an API for contract creation and data synchronization; SAML single sign on on the Pro plan; and an MCP server that lets Claude, ChatGPT, Cursor and other assistants draft from templates, file PDFs and start signature runs, with permissions following the user's account and actions written to the audit trail.
Slack and Microsoft Teams are listed as coming soon. Each integration has its own page, but no help center article, API reference or field mapping guide is published, so an implementer cannot see which objects move in which direction or what must be configured. Document management systems such as SharePoint, iManage or NetDocuments are not among the named integrations. A sales led company can connect its CRM; a legal team will need the API.
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.
Hosting and AI processing locations are published, and the tenancy model is not. The security and GDPR page says the platform is hosted on AWS with all contract data stored on servers in the European Economic Area, and that all AI processing happens within the EEA using Microsoft Azure OpenAI Service and Mistral AI. The subprocessor list, valid since 22 July 2026, refines that. It places the fynk app on AWS in the EEA, the fynk AI service on Hetzner in Germany, Azure AI Services and Mistral in the EEA, and Google Cloud in the EEA for simple and advanced electronic signatures.
Three processors, for SMS and WhatsApp delivery, support and error reporting, have occasional processing in the United States. Storage and processing locations are therefore given separately. Nothing states whether customers share infrastructure or are separated at the database level, and fynk is offered as one hosted service with no private cloud, region choice or on premises option.
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.
ISO 27001 is stated with its certification body, and no scope, date or certificate can be reached on the readable pages. The security and GDPR page says fynk's information security management system is ISO 27001 certified by DQS, accredited by the German accreditation body DAkkS, and that the hosting servers are themselves ISO/IEC 27001 certified. It also states a 99.5 percent annual availability target. No scope statement, certificate number, issue or expiry date, or route to obtain the certificate is published on that page, the pricing page or the legal pages, and no SOC 2 or other attestation is claimed.
A trust center exists at trust.fynk.com; its contents were not read on the date shown, so whether the certificate and any reports are open, available on request or held behind a sales conversation is not recorded here. Naming the certification body is more than a badge and lets a buyer check the certificate with DQS directly. A buyer has the standard, the body that issued it and a portal to ask through.
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.
The model providers and where they run are named, and the models and any notice of change are not. The security and GDPR page names Microsoft Azure OpenAI Service and Mistral AI as the AI providers and says all AI processing happens in the European Economic Area. The subprocessor list, valid since 22 July 2026, names Microsoft Ireland Operations for Azure AI Services in the EEA, Mistral AI for OCR in the AI features in the EEA, and Hetzner Online in Germany for hosting the fynk AI service, which tells a buyer that fynk runs its own AI service layer and calls the providers from there.
The list is dated and earlier versions from 27 February and 12 May 2026 are archived on the page, so changes can be traced after the fact. Not published is which models are used, such as which GPT model through Azure or which Mistral model, whether any model is fynk's own, and any commitment to tell customers before a provider or model changes. A legal team can tell its own clients which companies process its contracts with AI and in which region.
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.
Prices are published for every plan except the top one. The pricing page lists a free plan for one user with 50 AI credits a month, Essential at 19 euros per user a month with unlimited users and 150 AI credits, Business at 39 euros per user a month with 250 AI credits, and Pro at custom pricing with a custom credit allowance, a dedicated success manager, individual training, SAML single sign on and priority support.
Annual billing saves 120 euros per user a year against monthly billing, paid plans have a fourteen day trial without a credit card, all plans include unlimited documents under fair use, and electronic signatures carry no envelope fee. What is not published is the Pro price, the cost of extra AI credits, and whether onboarding or data migration costs extra. A small company can price fynk exactly before speaking to anyone; a company that needs single sign on or more AI than 250 credits a month has to ask.
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.
Who fynk serves is described with substance by team and by industry, and the boundaries are left open. Eight team pages cover legal, finance, operations, sales, IT, procurement, HR and founders and owners, and nine industry pages cover finance and banking, SaaS and IT, manufacturing, retail, energy, media and creative, sports, construction and ecommerce. Section II.3 of the terms limits the service to business customers.
The legal team page addresses in house legal work, contract visibility, approvals and risk surfacing, and the customer stories range from a solo founder handling all contracting to Conrad Electronic. What is not stated is where the product stops: which languages and governing laws its review and drafting are built for, whether law firms are a supported segment, and which contract types the AI analysis has been tuned on.
Electronic signatures are offered at the three eIDAS levels, which places the product in the European regulatory frame, and the published prices are in euros. A company team can see itself on the site; a legal team cannot tell from the site which of its contracts the AI handles well.
7 public documents
The public pages on file for fynk, with the recorded signals each one supports and the date it was last read. Open any of them and check the reading against the record.
-
Client Data in Training, Prompt and Output Retention, Third Party Request and Subpoena Notice and 1 more
Read Oct 2, 2026
-
fynk.com/en/features/ai-review3 signals
Primary Law Corpus Provenance, Good Law Verification, Refusal and Uncertainty Behavior
Read Oct 2, 2026
-
Outside Counsel Guideline Readiness
Read Oct 2, 2026
-
fynk.com/en1 signal
Billing and Fee Posture
Read Oct 2, 2026
-
fynk.com/en/integrations1 signal
Court Disclosure Support
Read Oct 2, 2026
-
fynk.com/en/security-and-gdpr1 signal
Ethical Walls and Matter Segregation
Read Oct 2, 2026
-
Fabricated Citation Record
Read Oct 2, 2026
€19 per user/monthEUR, as published, never converted
- fynk shows its prices on its website.
- There is a free plan for one person.
- Paid plans cost 19 or 39 euros for each user every month.
- Each plan includes a set number of AI uses a month.
- The biggest plan is priced on request.
Per user per month. Free: one user, 50 AI credits a month. Essential 19 euros: unlimited users, 150 AI credits. Business 39 euros: unlimited users, 250 AI credits, more integrations. Pro: custom price and credits, dedicated success manager, individual training, SAML single sign on, priority support. Annual billing saves 120 euros per user a year against monthly billing. All plans: unlimited documents under fair use, simple, advanced and qualified electronic signatures with no envelope fee, templates, approval workflows, audit trails. Fourteen day trial on paid plans without a credit card.
Implementation: Not stated.
Confidentiality and data terms: Not applicable. A data processing agreement is published on the legal page.
Note: The page does not say whether 19 and 39 euros are the annual or the monthly rate, only that annual billing saves 120 euros per user a year. The Pro price and the cost of extra AI credits are not published.
Legal Signals
What each signal meansA signal records what public sources say on the date shown. It is not a grade and it is not a recommendation. Where a signal reads Not addressed, it means the index did not locate the material in public sources on that date, which is a statement about disclosure rather than about the product.
Client Data in Training
Can material a lawyer puts into this product be used to train a model?
The published agreement expressly reserves a right to train on customer content, with no opt out located. Any de identification, anonymization or aggregation qualifier is recorded in the summary.
The agreement permits training on customer data once it has been anonymized. Section VII.3 of the terms for paid plans (version 2.3, 27 August 2026) grants an unlimited right to process customer data to improve the user experience and give feedback on AI functions, and bars passing it to third parties or training AI models on it unless anonymized first. The security and GDPR page says data is never used to train AI models; the agreement is narrower than that statement.
Prompt and Output Retention
How long does the product keep what a lawyer typed, and can that be set to zero?
Retention is acknowledged in public materials with no stated period.
Retention is tied to the life of the contract and no period is published for prompts or AI outputs. Section IX of the terms gives three months after termination to export contracts, after which the account and its data are deleted. Nothing on the security page, the terms or the subprocessor list says how long prompts and outputs are kept at the AI service or the model providers during the subscription.
Ethical Walls and Matter Segregation
Does retrieval respect the firm’s ethical walls, or can the model read across them?
No located public material addresses walls or matter level segregation.
No located public material describes how contracts are walled off between teams inside a customer or between customers. The security page limits fynk staff access to event related access by authorized employees, and the MCP integration page says permissions follow user accounts, without describing the permission model. Checked the security and GDPR, legal team, integrations, pricing and features pages and the terms on 2 October 2026.
Third Party Request and Subpoena Notice
If someone subpoenas the vendor for a firm’s data, does the firm hear about it first?
Terms commit to notice where lawfully permitted. No transparency report located.
The terms commit to telling the other party without delay when the law, a court order or an administrative decision requires disclosure. Section XI.3(e) also limits the disclosure to what is necessary and applies where no appeal is possible. No transparency report was located.
Primary Law Corpus Provenance
Where does the law in this product come from, and does the vendor have the right to use it?
No located public material identifies the corpus behind the product’s answers.
The product works on the customer's own contracts and the rules the customer sets, and does not retrieve primary law; no outside corpus behind its answers is identified. Checked the features, AI review and legal team pages on 2 October 2026.
Good Law Verification
Does the product tell you when the authority it just cited has been overruled?
No located public material addresses whether authority is checked for subsequent history.
The product does not cite case law or statutes, so no check of subsequent history applies, and none is described. Its review flags link to clauses in the contract under review. Checked the features, AI review and legal team pages on 2 October 2026.
Refusal and Uncertainty Behavior
What does the product do when the answer is not in the corpus?
No located public material addresses what the product does when it cannot ground an answer.
No located material describes what the AI does when it cannot support a finding. AI review marks each rule green, orange for partial or unclear compliance, or red, and the terms say AI output may include hallucinations; neither describes an abstention path or a confidence signal. Checked the AI review, features and legal team pages and the terms on 2 October 2026.
Fabricated Citation Record
Does a public court record exist addressing fabricated or hallucinated legal citations in output from this product?
No court order, opinion or disciplinary record addressing fabricated or hallucinated legal citations produced by this product has been located as of the date shown. This is a statement about the public record on that one subject, not a finding about the product, and this signal is not a litigation history.
No court order, opinion or disciplinary record naming fynk as the source of fabricated authority was located as of 2 October 2026. The AI Hallucination Cases database maintained by Damien Charlotin was searched on the name and returned no cases. This is a statement about the public record, not a finding about the product.
Bar Guidance Alignment
Has the vendor engaged in public with the ethics opinions its buyers are bound by?
Public materials refer to professional responsibility in general terms without naming guidance.
A no legal advice clause is published and no professional guidance is named. Section IV.8 of the terms for paid plans records that the customer knows the output is not legal advice; no bar, regulator or ethics guidance is referred to on the pages read.
Billing and Fee Posture
Does the vendor address what happens to the bill when the work takes an hour instead of six?
The product does not touch a fee between a lawyer and a client. It operates before an engagement exists, or it is bought by a team that bills no client for the work. Savings claims aimed at the buyer’s own cost are recorded in the summary and do not make the row a savings claim, because no client bill is in the loop.
The platform is sold to businesses for their own legal, sales, finance and operations teams, which bill no client for the work. The home page's 80 percent time saved and faster contract creation are savings to the buyer's own cost. No law firm segment or law firm customer was located.
Outside Counsel Guideline Readiness
Can a firm get this vendor through a client’s AI clause without a bespoke negotiation?
A subprocessor and model provider list plus client facing disclosure material is published or available without an agreement in place.
A dated subprocessor list naming the AI providers, a data processing agreement and a security page are all published without an agreement in place. The list, valid since 22 July 2026 with earlier versions archived, names each processor with purpose and location, including Azure AI Services, Mistral and Hetzner for the AI service; the security and GDPR page states where AI processing happens and links the list, the DPA and the terms.
Court Disclosure Support
If a judge’s standing order requires an AI disclosure, can the product produce one?
Some elements of the record are available, short of a document level export.
Some elements of a record exist in a product built for company contract work rather than filings. Audit trails and version control are in every plan, AI review flags link to the clauses they rest on, and actions taken through the MCP server are written to the audit trail. Nothing published offers a per document export of which model produced an output and what a person checked, and court disclosure is not addressed.