Dixa

Summary of Important Changes

1. Document structure and incorporation into the contract

• The document title and introductory text have been substantially reformatted.
• The Privacy Policy is now expressly stated to be incorporated into and subject to the Dixa Terms of Service, rather than merely referring generally to “Dixa Terms.”
• The definition of the “Agreement” is clarified as the Dixa Terms of Service and associated procedures.
• Additional website navigation links and policy titles have been inserted, including the Legal Hub, GDPR information, Data Processing Agreement, Data Security, Early Access Terms, and Referral Program. These additions appear primarily editorial, but they could create uncertainty if those linked documents are intended to form part of the contractual framework.

Risk: Customers should verify which linked policies and procedures are legally binding and whether they can be amended independently.

2. Definitions and scope

• Definitions of “Dixa,” “us,” “we,” “Website,” “Dixa Services,” and “Customer Data” have been cleaned up and standardized.
• References to Personal Data now use more consistent terminology.
• The wording continues to state that Personal Data contained in Customer Data is subject to the Customer’s own privacy policy and that data-subject requests should generally be directed to the Customer.

Risk: The allocation of responsibility between Dixa and the Customer remains important, particularly where the Customer determines the purposes of processing and Dixa acts as a service provider or processor.

3. Data use and analytics

• The policy continues to state that information may be used to analyze usage trends and preferences, optimize the Dixa Services, monitor traffic and user movements, and gather demographic information.
• No express new language authorizes Dixa to use Customer Data for product development, generalized analytics, machine learning, artificial intelligence, or model training.

AI training: The supplied diff does not identify any change expressly permitting or restricting the use of Customer Data to train AI models. It also does not add safeguards addressing model training, de-identification, retention of training data, or use of customer prompts, conversations, or outputs. Customers should not assume that AI-training rights are excluded merely because they are not visible in this excerpt; those rights may appear in the Terms of Service, Data Processing Agreement, or another product-specific document.

4. International hosting and subcontractors

• The policy retains language stating that Dixa Services are hosted in the United States or EU and that EU customers’ databases will be hosted in an EU data center.
• It continues to disclaim responsibility for Customer Data stored on Dixa or subcontractor servers at the Customer’s discretion.

Risk: Customers should confirm applicable international-transfer mechanisms, subprocessors, and security obligations in the DPA.

5. Other changes

• Minor grammatical and punctuation changes were made.
• The contact email for privacy concerns is clarified as privacy@dixa.com.
• “Do Not Track” terminology and opt-out wording are standardized.

Summary of Important Changes

1. Expanded and reorganized terms

• The terms now incorporate or link to a broader Legal Hub, including the Privacy Policy, Cookie Policy, GDPR materials, Data Processing Agreement, security information, service-level terms, and other policies.
• A new defined term, “EAT” (Early Access Terms), is introduced, together with a referral-program reference.
• The agreement now expressly includes the Order Form and documents referenced in it, increasing the importance of reviewing those documents for conflicting or additional obligations.

2. New and expanded product definitions

The definitions section has been substantially expanded. Important additions include:

• “Add-On”, expressly including AI-related products such as Mim AI Co-Pilot and AI Quality Assurance.
• “Minimum Commitment”, which includes agent seats, add-ons, telecommunications packages, and usage-based services such as conversations handled by AI agents.
• “Overage”, which expressly includes exceeding limits for AI analysis, API calls, agents, voice minutes, and messages.
• “Collaboration User”, replacing the former “View-Only User” terminology and limiting those users to internal, non-customer-facing activities.
• Expanded definitions of Customer Data, Content, Services, Voice Communication, SMS, MMS, and telephony services.

These changes may broaden the services covered by the agreement and clarify that AI usage can create binding minimum commitments and additional usage charges.

3. AI and use of customer data

The most important data-use provision states that Dixa may:

• Aggregate, collect, and analyze data and information relating to the provision, use, and performance of the Services, including traffic and telemetry information;
• Use that information during and after the agreement to develop and improve the Services and other Dixa offerings; and
• Disclose it only in an aggregated and anonymized format that does not identify the Customer or an individual.

Risk assessment

• This is an express permission for post-termination use of service-related data for product development and improvement.
• The clause is limited to data that “does not permit identification” of an individual and disclosure that does not identify the Customer. However, it does not expressly state that Customer Content, prompts, transcripts, or outputs are excluded from AI training or model development.
• It also does not clearly distinguish between ordinary analytics, AI model training, service improvement, and development of unrelated Dixa offerings.
• Customers should seek clarification or a contractual carve-out confirming that Customer Data and Content will not be used to train general-purpose or shared AI models unless separately authorized.

4. Commercial and operational risks

• Subscriptions generally cannot be downgraded or cancelled during the current term, and minimum commitments remain billable even if usage falls.
• Dixa may charge Overage at then-current rates, including for AI analysis and AI-agent usage.
• Fees automatically increase by a compounded 5% annual indexation, without separate notice.
• Dixa may change pricing, services, Documentation, and terms, subject to limited notice requirements.
• Dixa may suspend access for breach or non-payment, and liability is capped at fees paid during the preceding six months.

5. Other notable changes

• Dixa receives broad rights to use the Customer’s name and logo in marketing and may request a joint public announcement.
• Services and infrastructure may be located outside the EU, subject to an approved transfer mechanism.
• Customers are broadly responsible for their data’s legality and may bear third-party royalties, fines, and settlements.

Summary of Important Changes

Overall Changes

• The Terms are presented as updated 28 May 2025 and are substantially reorganized and expanded.
• The agreement now contains more detailed definitions, ordering/payment rules, service restrictions, data provisions, AI-related products, and telecommunications terms.
• The revised Terms apply not only to signed Order Forms but also to online registration, activation, and use of the Services.

AI and Customer Data

AI products and usage

• “Add-On” now expressly includes AI products such as Mim AI Co-Pilot, AI Quality Assurance, and Advanced Insights.
• “Overage” expressly includes:
• Use of AI tools or services beyond purchased limits;
• Usage by AI agents; and
• Other usage-based AI features identified in an Order Form.
• AI-related usage may therefore generate additional, non-refundable charges, and minimum commitments may continue regardless of actual usage.

Use of data to improve services and possible AI training

A new or materially clarified provision allows Dixa to:

• Aggregate, collect, and analyse data and information relating to the provision, use, and performance of the Services, including traffic and telemetry data;
• Use that information during and after the Agreement to develop and improve the Services and other Dixa offerings; and
• Disclose it only in aggregated and anonymised form that does not identify the Customer or any individual.

The clause does not expressly say “train AI models” or expressly authorize training on identifiable Customer Data. However, “develop and improve” the Services and “other Dixa offerings” is broad enough that Dixa may argue it covers development or training of AI systems, provided the data is aggregated and anonymised.

Key data risks

• “Customer Data” includes information uploaded by users and Content supplied by the Customer’s end-users, potentially including communications and personal data.
• The Terms do not clearly define the anonymisation standard, prohibit re-identification, or state whether prompts, transcripts, outputs, or other AI inputs are retained for model development.
• The permission is not limited to the specific Customer, Service, or purpose and continues after termination.
• Customers requiring a strict prohibition on AI training or secondary use should seek an express contractual restriction and confirm the position in the DPA, product terms, and AI documentation.

Other Important Commercial and Legal Changes

• Automatic renewal: 12-month renewals unless non-renewal notice is given at least three months before expiry.
• Pricing: Automatic compounded 5% annual indexation, plus possible immediate changes to published list and overage prices.
• Minimum commitments: Commitments generally cannot be reduced during a term, even if usage falls.
• Overages: Charged at then-current rates, generally without discounts, and are non-cancellable and non-refundable.
• Unilateral amendments: Dixa may change the Terms with 30 days’ notice, with continued use treated as acceptance.
• Liability: Dixa’s aggregate liability is capped at Fees paid during the preceding six months, with broad exclusions for data loss and indirect loss.
• Data loss: The stated remedy is limited to commercially reasonable restoration from Dixa’s latest backup.
• Marketing: Customer name and logo may be used for promotional purposes, and Dixa may request a joint public announcement.

Executive Summary

The diff is largely a formatting and website-navigation update rather than a substantive rewrite of privacy rights or data-use permissions. However, several points may affect interpretation or should be checked against the full policy.

Important Changes

1. Policy title, navigation, and date information

• The opening text now includes a broader Legal Hub navigation menu, including links to:
• Terms of Service
• GDPR
• Data Processing Agreement
• Data Security
• Applicant Privacy Notice
• Accessibility Statement
• Early Access Terms
• Referral Program
• The document now expressly identifies itself as “Privacy Policy” and states: “Last updated: 25 Novembre 2021.”
• This appears primarily to be website formatting or header content, not a change to substantive privacy terms.
• The date is unusual because it remains in French (“Novembre”), which could create uncertainty about whether the document has actually been updated.

2. Incorporation into the Terms of Service

• The policy now states that it is incorporated into and subject to the Dixa Terms of Service, rather than referring more generally to “Dixa Terms.”
• The definition of “Agreement” is clarified to mean the Terms of Service and associated procedures.
• This may strengthen the contractual relationship between the Privacy Policy and the Terms of Service. Customers should review the Terms of Service and any referenced procedures for additional data-use rights, liability limits, or conflicts.

3. Clarification of defined terms

The definitions of:

• “Dixa,” “us,” and “we”
• “Website”
• “Dixa Services”
• “Customer Data”
• “Personal Data”

are reformatted and grammatically standardized. No clear substantive expansion of the types of data covered is shown in the diff.

4. Data-use language

The policy continues to state that information may be used to:

• Analyze usage trends and preferences
• Administer or optimize the Dixa Services
• Monitor usage or traffic patterns, including user movements
• Gather demographic information
• Understand customers’ and visitors’ interests, habits, and usage patterns

These purposes are broad and could support analytics, product improvement, and service optimization. The diff does not clearly expand these purposes, but they remain potentially important from a customer-risk perspective.

5. Customer Data and privacy requests

• The policy continues to state that Personal Data forming part of Customer Data is subject to the customer’s own privacy policy.
• Requests for access, correction, or deletion should generally be directed through the customer, because Dixa may not have a direct relationship with the customer’s end users.
• “Opt-out” wording is standardized but no new opt-out right is apparent.

6. Security, cookies, and hosting

• Minor wording and punctuation changes were made concerning security breaches, cookies, and “Do Not Track” signals.
• The policy continues to state that services are hosted in the United States or EU, with EU customer databases hosted in an EU data center, subject to the surrounding qualifications.
• No clear change to international-transfer or subcontractor responsibility provisions is shown.

AI Model Training

• No express provision authorizing or prohibiting the use of Customer Data to train AI or machine-learning models has been added or removed in this diff.
• The broad “analyze,” “optimize,” and “understand usage patterns” language could potentially be argued to cover product analytics or improvement, but it does not expressly mention AI training, model training, model development, or use of customer content for machine learning.
• Customers seeking certainty should request a written clarification or review the Terms of Service, Data Processing Agreement, and Data Security materials for separate AI-training provisions.

Summary of Important Changes and Risks

1. Major expansion and restructuring of the Terms

The diff appears to replace a short version of the Terms with a substantially expanded agreement. It adds:

• A detailed definitions section;
• Rules for account access, permitted users, integrations, and prohibited conduct;
• Consultancy and onboarding terms;
• Service levels and support;
• Fees, overages, renewals, suspension, and termination;
• Intellectual-property provisions;
• Customer-data, data-protection, and confidentiality provisions;
• Warranties, liability limits, indemnities, marketing rights, and governing-law provisions.

Customers should review the complete replacement carefully, as many obligations and limitations are newly stated rather than merely clarified.

2. AI-related products, usage, and charges

The Terms now expressly define “Add-On” products, including:

• Mim AI Co-Pilot;
• AI Quality Assurance;
• Advanced Insights; and
• Usage involving AI agents.

The definition of “Minimum Commitment” expressly includes usage-based subscriptions, such as conversations handled by AI agents. “Overage” now includes analyzing more data with AI tools or services than the purchased plan permits. Excess usage may be charged at then-current rates, without discount, and is generally non-cancellable and non-refundable.

Risk: Customers may incur additional AI-related charges based on usage or analysis volumes, while pricing and applicable thresholds may be located in an Order Form, platform, or separate price list. Customers should confirm how AI usage is measured and whether administrators can activate chargeable features.

3. Use of Customer Data to improve services and potentially train AI

The Terms grant Dixa a continuing right, during and after the agreement, to:

• Aggregate, collect, and analyze data and information relating to the provision, use, and performance of the Services, including traffic and telemetry data;
• Use that information to develop and improve the Services and other Dixa offerings; and
• Disclose it only in aggregated and anonymized form that does not identify the Customer or an individual.

The clause does not expressly say that Dixa may use identifiable Customer Data, conversations, or personal data to train AI models. However, “data and other information relating to” service use is broad, and “develop and improve” could potentially include developing or training machine-learning or AI systems, provided Dixa considers the information non-identifying.

Key risks:

• No explicit exclusion of Customer Content, transcripts, prompts, or outputs from AI training or model development;
• No detailed standard for anonymization, re-identification risk, retention, or deletion;
• The right continues after termination;
• The clause says Dixa obtains rights to the aggregated and anonymized data;
• Customers requiring a strict “no training” commitment may need a separate contractual restriction or DPA amendment.

4. Customer data and privacy responsibilities

The Customer retains ownership of Customer Data but is solely responsible for its legality, accuracy, and third-party rights. The Customer is designated as data controller, while Dixa’s standard DPA applies unless a separate DPA takes precedence.

Special-category personal data is prohibited unless specifically agreed in the DPA. Services are stated not to be HIPAA- or PCI-DSS-compliant. Data may be processed outside the EU, subject to an approved transfer mechanism.

5. Commercial and operational risks

Notable customer-unfavorable provisions include:

• Automatic 12-month renewals unless notice is given at least three months before expiry;
• Annual compounded 5% fee indexation;
• Minimum commitments that cannot be reduced during the term;
• Immediate suspension for payment problems;
• Broad rights to change Services and amend the Terms;
• A liability cap generally limited to six months’ subscription fees;
• Broad customer indemnification obligations;
• Permission to use the Customer’s name and logo in marketing;
• Exclusive Danish or Delaware courts, depending on the contracting Dixa entity.

Summary of Important Changes

1. AI and use of Customer Data

• New AI-related products are expressly included. “Add-On” now includes products such as Mim AI Co-Pilot, AI Quality Assurance and Advanced Insights.
• AI usage may generate additional charges. “Minimum Commitment” and “Overage” now expressly cover usage of AI tools, including analysing data with AI services and conversations handled by AI agents. Excess usage may be billed at then-current rates, without discounts, and is generally non-cancellable and non-refundable.
• New data-use right for Dixa. Dixa may “aggregate, collect and analyse” data and information relating to the provision, use and performance of the Services, including traffic and telemetry data, provided this is done in a manner that does not permit identification of an individual.
• Dixa may use this aggregated and anonymised information, during and after the Agreement, to:

1. develop and improve the Services; and

2. develop and improve other Dixa offerings.

• Dixa may disclose the information only in an aggregated and anonymised format that does not identify the Customer or any individual, and obtains rights in that data.

AI-training risk

The Terms do not expressly say that identifiable Customer Data or Customer content may be used to train general-purpose AI models. However, the broad right to analyse data to “develop and improve” the Services and “other Dixa offerings” could potentially include machine-learning or AI model development, provided Dixa considers the resulting data aggregated and anonymised.

Key uncertainties and risks include:

• “Aggregated” and “anonymised” are not defined, and the clause does not specify the anonymisation standard or re-identification safeguards.
• The restriction is framed around identifying an individual; it may not clearly prevent identification of the Customer, a business, or a particular account in every circumstance.
• There is no express customer opt-out, deletion right, audit right, or restriction against using derived insights, prompts, transcripts, or model outputs for AI improvement.
• The clause applies after termination.

Customers handling confidential, regulated, or commercially sensitive data should seek a specific contractual restriction on AI training and clarify whether AI add-ons process data for model training.

2. Commercial commitments and pricing

• Minimum commitments now expressly include agent seats, add-ons, telephony packages and AI usage.
• A 5% annual compounded fee increase applies automatically without separate notice.
• Dixa may change list and overage prices immediately by posting them in the Platform.
• Subscriptions generally cannot be downgraded or cancelled during the current term, and the Agreement automatically renews for 12-month periods unless notice is given at least three months before expiry.

3. Liability and data protection

• Dixa’s aggregate liability is capped at fees paid in the preceding six months.
• The customer’s remedy for data loss is limited to commercially reasonable restoration from Dixa’s latest backup.
• The customer remains responsible for the legality and accuracy of Customer Data and generally may not process sensitive personal data unless agreed in the data processing agreement.
• Dixa may use non-EU/EEA providers subject to an approved transfer mechanism.

Summary

• The diff states only: “Added approximately 23 words to the document.”
• The actual added language is not provided, so it is not possible to determine:
• What contractual rights or obligations changed;
• Whether liability, confidentiality, data protection, or termination provisions were affected;
• Whether the customer’s data may be used for AI training or model improvement; or
• Whether any new risks were introduced.

AI Training and Data Use

• No specific AI-related language is identifiable from the diff provided.
• In particular, the diff does not show whether the agreement:
• Permits the provider to use customer data, prompts, outputs, or usage data to train or fine-tune AI models;
• Limits such use to de-identified, aggregated, or service-improvement purposes;
• Requires customer consent or provides an opt-out;
• Allows human review of customer data; or
• Restricts retention, disclosure, or transfer of data to model-training systems.

Risk Assessment

Because the 23 added words are not shown, the legal significance cannot be assessed reliably. The added wording should be supplied in full, including any surrounding sentence or paragraph, before signing or relying on this analysis.

Structured Summary of Important Changes

1. New and expanded definitions

The revised Terms add an extensive definitions section, including:

• “Add-On”, expressly covering AI products such as Mim AI Co-Pilot and AI Quality Assurance.
• “Customer Data”, broadly covering information supplied by the Customer, uploaded by users, and content provided by the Customer’s end-users.
• “Minimum Commitment”, which may include agent seats, add-ons, telecommunications packages, and usage-based products such as conversations handled by AI agents.
• “Overage”, expressly including excess use of AI tools and services, as well as excess agents, minutes, messages, and API calls.
• “Collaboration User”, replacing the prior “View-Only User” terminology and limiting those users to internal, non-customer-facing activities.

Risk: AI functionality and AI-related usage charges are now expressly incorporated into the contractual pricing and commitment structure. Customers may remain liable for minimum AI-related commitments and uncapped or changing overage charges, even if actual use is lower than expected.

2. AI data use and model training

The key data-use provision states that Dixa may:

• Aggregate, collect, and analyse data and information relating to the provision, use, and performance of the Services, including traffic and telemetry data.
• Use that information during and after the Agreement to develop and improve the Services and other Dixa offerings.
• Disclose it only in an aggregated and anonymised format that does not identify the Customer or any individual.

The clause does not expressly say “training AI models”, nor does it expressly prohibit Dixa from using anonymised or aggregated Customer-related data to train, fine-tune, evaluate, or improve AI models. It also does not clearly distinguish between ordinary service telemetry and Customer Data used by AI features.

Risk: Although the clause contains safeguards against identifying the Customer or individuals, it gives Dixa a continuing, post-termination right to use aggregated and anonymised information for product development. Customers seeking a strict “no use of Customer Data for AI training” commitment should request clearer language addressing model training, prompts, outputs, recordings, transcripts, and human review.

3. Data protection and confidentiality

The revision:

• Confirms that the Customer is the data controller and incorporates Dixa’s standard Data Processing Agreement.
• Restricts use of the Services for special-category/sensitive personal data unless specifically agreed in the DPA.
• Permits processing using equipment outside the EU, subject to an approved transfer mechanism such as Standard Contractual Clauses.
• Treats Customer Data and customer-related business information as Confidential Information.

Risk: The Terms place substantial responsibility on the Customer for the legality, accuracy, and third-party rights relating to Customer Data, including potential royalties, fines, and settlements.

4. Commercial and operational changes

The revised Terms add or clarify:

• A 5% annual compounded fee indexation, automatically applied without separate notice.
• Minimum commitments that generally cannot be reduced during the subscription term.
• Broad overage billing at then-current, non-discountable rates.
• Automatic 12-month renewals unless notice is given at least three months before expiry.
• Dixa’s right to modify Services and Documentation.
• Broad suspension rights for violations or non-payment.

Risk: Customers face increased price escalation, renewal, overage, and service-change exposure, with limited cancellation or refund rights.

Summary of the Available Change

Scope of Change

• The diff states that approximately 9,125 words were removed from the document.
• No replacement language or details of the deleted provisions are provided.

AI Training and Customer Data

• It is not possible to determine from the provided diff whether the agreement now permits, restricts, or prohibits using customer data to train AI models.
• The deleted material may have included provisions addressing:
• Use of customer content, prompts, inputs, outputs, or personal data for model training or improvement;
• Whether training is conducted on an aggregated, de-identified, or identifiable basis;
• Opt-in or opt-out rights;
• Retention and deletion of data;
• Human review or use of data for service improvement;
• Ownership of customer data and AI-generated outputs; and
• Confidentiality, security, and regulatory obligations.

Key Legal Risks Created by the Deletion

• Loss of protections: Removing data-use, confidentiality, security, or purpose-limitation provisions could broaden the provider’s practical ability to use customer information.
• Ambiguity: If the deleted language was not replaced, the agreement may no longer clearly state whether customer data can be used for AI training.
• Reduced customer rights: Deleted opt-out, deletion, audit, notice, or consent provisions could weaken the customer’s control over its data.
• Compliance uncertainty: Removing privacy and data-processing terms could create uncertainty under applicable privacy, confidentiality, sector-specific, or contractual requirements.
• Conflicting documents: The operative position may now depend on incorporated policies, product terms, or privacy notices that were not included in the diff.

Conclusion

The only confirmed change is a substantial deletion of approximately 9,125 words. A substantive legal review cannot reliably identify the actual changes or AI-training implications without the deleted text, the revised agreement, or a full redline. The deletion should be treated as high-risk until it is confirmed that customer-data use and AI-training provisions remain adequately addressed.

Summary of Important Changes

1. Substantial expansion and restructuring of the Terms

The diff replaces the prior short introductory text with a comprehensive set of Terms of Service and definitions. It adds or clarifies the contractual framework, including:

• A detailed list of related legal documents, including the Privacy Policy, Cookie Policy, GDPR materials, Data Processing Agreement, Data Security information, Service-Level Agreement, and Early Access Terms.
• Rules for customers accepting the Terms on behalf of an organisation.
• A comprehensive definitions section.
• Detailed provisions on services, fees, renewals, intellectual property, data protection, confidentiality, liability, indemnities, marketing, amendments, subcontractors, and governing law.

These provisions substantially increase the customer’s contractual obligations and Dixa’s ability to modify, suspend, charge for, or terminate services.

2. AI-related products, usage and charging

The revised definitions expressly add AI-related products and usage:

• “Add-On” includes AI products such as Mim AI Co-Pilot and AI Quality Assurance.
• “Minimum Commitment” may include usage-based subscriptions, including conversations handled by AI agents.
• “Overage” includes analysing more data with AI tools or services than is included in the applicable plan.
• AI Add-Ons may be priced by Agent seat, with excess usage charged at then-current Overage rates.

Risks

• Customers may be committed to minimum AI-related volumes or seats even if actual usage is lower.
• AI-related overage charges may be imposed at current rates, without discounts, and are non-cancellable and non-refundable once incurred.
• The Terms do not state specific AI performance standards, accuracy requirements, human-review obligations, or liability protections for AI outputs.

3. Use of data to train or improve AI models

The revised confidentiality clause gives Dixa the right to:

> aggregate, collect and analyse data and other information relating to the provision, use and performance of the Services, including traffic and telemetric information

Dixa may use this information during and after the Agreement to:

• Develop and improve the Services; and
• Develop and improve other Dixa offerings.

Dixa may disclose the information only in aggregated and anonymised form that does not identify the Customer or any individual.

AI-training implications

• There is no express statement that Customer Data, message content, recordings, or other customer-provided content may be used to train general-purpose AI models.
• However, the broad right to analyse information to “develop and improve” the Services and other offerings could potentially include service analytics, model evaluation, or AI-system improvement, depending on how Dixa interprets the clause.
• The clause is limited to information that does not identify an individual, but it does not expressly address re-identification risk, pseudonymised data, prompts, outputs, call recordings, or model-training datasets.
• Customers seeking to prohibit use of their data for AI training should require an express contractual restriction, preferably in the Order Form or Data Processing Agreement.

4. Commercial and operational risks

• Automatic annual 5% compounded fee indexation, without separate notice.
• Dixa may immediately change list and Overage prices by posting them on the Platform.
• Automatic 12-month renewals unless notice is given at least three months before expiry.
• Minimum Commitments cannot generally be reduced during the Subscription Term.
• Dixa may suspend accounts for breach or non-payment.
• Dixa may modify the Services and Documentation at its discretion.
• Liability is capped at Fees paid during the preceding six months, with broad exclusions for indirect loss and data loss.
• Customers must indemnify Dixa for claims arising from their use of the Services.
• Dixa may use the Customer’s name and logo in marketing and may request a joint public announcement.

Summary

The provided diff only states:

> “Added approximately 23 words to the document”

It does not include the actual added, deleted, or replaced legal text.

Important Changes

• Cannot be determined: The substantive effect of the amendment cannot be assessed without the 23 added words and the surrounding contract language.
• No identifiable legal risks: No specific changes to obligations, liability, confidentiality, intellectual property, data rights, termination, or other provisions are visible.

AI Training and Customer Data

• No determination possible: The diff does not reveal whether the customer’s data may be:
• Used to train, fine-tune, or improve AI models;
• Shared with affiliates, vendors, or third parties for model development;
• Retained or aggregated for training purposes;
• Used only in de-identified form; or
• Excluded from training unless the customer opts in.

Information Needed

Please provide the actual redlined text, including:

• Additions in {braces};
• Deletions in [brackets]; and
• Replacements in []{}

The surrounding language should also be included where necessary to determine how the amendment changes the customer’s rights and the provider’s permissions.

Summary of Important Changes

1. New Terms of Service structure and effective date

• The Terms are presented as updated Terms of Service dated 28 May 2025.
• The agreement now expressly incorporates the applicable Order Form and referenced documents as part of the binding agreement.
• The definitions and overall structure have been substantially expanded and reorganized.

2. AI products and AI-related charges

• “Add-On” is newly defined to include AI products such as:
• Mim AI Co-Pilot
• AI Quality Assurance
• Advanced Insights and other modules.
• “Overage” expressly includes:
• Usage exceeding purchased AI-analysis quantities or limits.
• Usage involving AI agents, where applicable.
• AI Add-Ons priced per Agent generally track the number of Agents committed under the Main Plan. Additional usage may be charged at then-current Overage rates.
• Overage charges are generally non-cancellable and non-refundable, and may be billed monthly without discounts.
• Customers may therefore face additional fees if AI usage or the number of users exceeds the quantities in the Order Form.

3. Customer data and AI model training

• “Customer Data” is defined broadly to include:
• Information supplied by the Customer.
• Data uploaded by Permitted Users.
• Content provided by the Customer’s end-users.
• The new confidentiality clause expressly allows Dixa to aggregate, collect and analyse data relating to the provision, use and performance of the Services, including traffic and telemetry information.
• Dixa may use this information during and after the agreement to:

1. Develop and improve the Services and other Dixa offerings; and

2. Disclose it only in aggregated and anonymised form that does not identify the Customer or any individual.

• The clause states that the data must not permit identification of any individual and gives Dixa rights in the resulting aggregated and anonymised data.

AI-training risk

• The wording does not expressly state that Customer Data or customer content will be used to train generative AI or machine-learning models.
• However, “develop and improve the Services and other Dixa offerings” is broad enough to potentially cover model development, testing, analytics, or improvement—particularly where AI Add-Ons are used.
• The clause does not specify:
• Whether data is excluded from training by default.
• Whether customers can opt out.
• What de-identification or anonymisation standards apply.
• How long data may be retained for these purposes.
• Whether human review or subcontractor/model-provider access is permitted.
• Customers handling confidential, regulated, or sensitive information should seek clarification or a contractual restriction before enabling AI features.

4. Data protection and sensitive information

• The Customer is expressly treated as the data controller, with Dixa’s standard Data Processing Agreement applying unless separately replaced.
• Use of special-category personal data under GDPR is prohibited unless specifically agreed in the DPA.
• The Services are stated not to be HIPAA- or PCI DSS-compliant.
• Services and equipment may be located outside the EU, subject to an approved transfer mechanism such as Standard Contractual Clauses.

5. Commercial and operational risks

• A 5% annual compounded fee indexation is introduced, applying automatically without separate notice.
• Minimum Commitments cannot generally be reduced during the Subscription Term, even if actual usage is lower.
• The agreement automatically renews for successive 12-month periods unless notice is given at least three months before expiry.
• Dixa may modify the Services and Documentation, and may change immediate Overage and list prices through the Platform.
• Dixa may suspend accounts for breach or non-payment and disclaims liability for third-party integrations and data loss beyond reasonable restoration efforts.

Summary

The provided diff only states: “Added approximately 23 words to the document.” It does not include the actual added language or identify where it appears.

AI Training and Data-Use Changes

• Cannot be determined from the provided diff.
• No specific wording is available showing whether customer data may be:
• Used to train, fine-tune, or improve AI models;
• Combined with other customers’ data;
• De-identified or anonymized before use;
• Reviewed by humans for model development;
• Retained after termination; or
• Shared with affiliates, vendors, or third parties.

Legal Risks

Because the 23 added words are not provided, it is not possible to assess whether they:

• Expand the provider’s license or rights in customer data;
• Create a new consent to AI training or model improvement;
• Permit use of confidential, personal, or regulated information;
• Change ownership or intellectual-property rights;
• Limit the customer’s ability to opt out;
• Reduce confidentiality, security, or deletion obligations; or
• Create broader disclaimers or liability protections for AI-related use.

Required Information

To perform a meaningful legal analysis, provide the actual text of the 23-word addition, including the surrounding clause if possible.

Summary

Scope of the Diff

The provided diff states only:

> “Added approximately 9125 words to the document”

It does not include the actual added contractual language or identify which provisions were changed.

Legal and Commercial Impact

The changes cannot be meaningfully analyzed without the text of the additions. In particular, it is not possible to determine whether the new language:

• Expands the provider’s rights to use customer data;
• Permits customer data to be used to train, fine-tune, evaluate, or improve AI models;
• Allows use of customer data for product development, analytics, benchmarking, or advertising;
• Distinguishes between personal data, confidential information, prompts, outputs, and de-identified or aggregated data;
• Changes ownership or licensing rights in customer data, inputs, outputs, or model improvements;
• Permits disclosure of customer data to affiliates, subprocessors, vendors, or third-party AI providers;
• Adds data-retention, deletion, security, or confidentiality exceptions;
• Changes the customer’s consent or opt-out rights;
• Limits the provider’s liability for AI-related data use or security incidents; or
• Applies the new terms retroactively to previously collected data.

AI Training Risk

No conclusion can be reached regarding AI-model training because the relevant provisions are not included. The most important language to review would include terms such as:

• “train,” “fine-tune,” “improve,” “develop,” or “enhance” models;
• “inputs,” “content,” “customer data,” or “usage data”;
• “de-identified,” “aggregated,” or “derived data”;
• “service improvement” or “product development”;
• “perpetual,” “irrevocable,” “worldwide,” or “royalty-free” licenses; and
• rights to share data with third parties or subprocessors.

Required Information

Please provide the actual redline text, including the approximately 9,125 added words, or at least the sections concerning:

1. Data use and licensing;

2. AI or machine-learning services;

3. Confidentiality and security;

4. Subprocessors and third-party providers;

5. Data retention and deletion;

6. Intellectual property; and

7. Liability and indemnification.

Without the substantive text, any assessment of legal risk would be speculative.

Summary of Changes

Scope of the Diff

• The diff states that approximately 9,125 words were removed from the document.
• No replacement or newly added language is provided.
• Because the deleted text is not identified, it is not possible to determine which specific rights, obligations, limitations, or protections were removed.

Potential Legal Significance

A deletion of this size could materially change the agreement, particularly if it removes provisions concerning:

• Confidentiality and data protection
• Ownership or licensing of customer data
• Service-provider obligations and security standards
• Liability limitations, indemnities, and warranties
• Termination and data-return/deletion rights
• Audit, compliance, or regulatory requirements
• Dispute resolution and governing law
• Changes to pricing, service levels, or renewal terms

The removal of previously stated protections may benefit the party retaining the remaining language, but the impact cannot be assessed without seeing the deleted provisions and the resulting contract.

AI Training and Customer Data

• The provided diff does not identify whether any language concerning AI training was added, removed, or changed.
• It is therefore unclear whether the revised agreement:
• Permits or prohibits using customer data to train, fine-tune, or improve AI models
• Limits use to aggregated, de-identified, or anonymized data
• Covers prompts, inputs, outputs, telemetry, or usage data
• Requires customer consent or provides an opt-out
• Allows use by affiliates, contractors, or third-party AI providers
• Establishes retention, deletion, security, or ownership rules for training data
• Addresses whether customer data may be used to train models shared with other customers

If the deleted text contained restrictions on AI training or data use, its removal could create a significant new risk by eliminating those restrictions. Conversely, if it contained a permission to use data for training, deleting it could narrow the provider’s rights.

Recommended Review

Obtain:

1. The full pre-change and post-change versions;

2. A redline identifying the actual deleted provisions; and

3. Any revised definitions of “Customer Data,” “Usage Data,” “Aggregated Data,” “AI,” or “Services.”

Until those materials are available, no definitive conclusion can be reached about the effect on customer data or AI-model training rights.

Summary

The provided diff does not include the actual amended contract language. It only states:

> “Added approximately 23 words to the document”

Accordingly, it is not possible to determine:

• What provisions were changed;
• Whether the change expands or limits either party’s rights or obligations;
• Whether customer data may be accessed, retained, shared, or disclosed;
• Whether customer data, prompts, outputs, or usage information may be used to train, fine-tune, evaluate, or improve AI models;
• Whether any consent, opt-out, confidentiality, security, ownership, or deletion rights were modified; or
• Whether the amendment creates new liability, indemnity, compliance, or regulatory risks.

AI Training Risk Review

No AI-training-related language appears in the supplied diff. However, the absence of text does not establish that the contract contains no such change; the actual 23 added words are required for review.

The most important terms to check for include:

• Training authorization: “use,” “process,” “analyze,” “improve,” “develop,” “train,” or “fine-tune” models;
• Data scope: customer content, prompts, outputs, metadata, usage data, or de-identified/aggregated data;
• Permission structure: whether use is automatic, optional, opt-in, or subject to an opt-out;
• Ownership and confidentiality: whether training use affects customer ownership or confidentiality protections;
• Retention and deletion: whether data is retained after termination or deletion requests; and
• Third-party access: whether vendors or model providers may receive or use the data.

Conclusion

No substantive legal or AI-data-use conclusions can be drawn from the placeholder description alone. Please provide the actual redlined text, including the added words and any surrounding sentence or clause, for a reliable risk analysis.

Summary of Important Changes

1. New or Expanded AI-Related Provisions

• AI products are expressly included as Add-Ons. The updated terms identify products such as Mim AI Co-Pilot and AI Quality Assurance as Add-Ons. This makes AI functionality part of the contractual product structure and potentially subject to separate subscription, usage and billing rules.
• AI usage may generate overage charges. “Overage” now expressly includes analysing more data with AI tools than the amount included in the applicable plan. AI usage can therefore create additional, non-refundable charges at Dixa’s then-current rates.
• No express right to use identifiable Customer Data for model training was added. The new data-use language states that Dixa may aggregate, collect and analyse data relating to the provision, use and performance of the Services, including traffic and telemetry information, only in a way that does not permit identification of any individual.
• Dixa may use this aggregated and anonymised information, during and after the contract term, to:
• develop and improve the Services; and
• develop and improve other Dixa offerings.
• Dixa may disclose the information only in an aggregated and anonymised form that does not identify the Customer or any individual, and may retain rights in that data.

AI training risk

Although the clause does not expressly say “train AI models,” “develop and improve” the Services and other offerings could potentially include training or fine-tuning AI systems. The wording is not explicit about:

• whether Customer Data, call recordings, messages or prompts may be used;
• whether de-identification occurs before AI processing;
• whether inputs or outputs are retained;
• whether data may improve models used by other customers; or
• whether the Customer can opt out.

The “no individual identification” restriction may not fully address confidential business information or re-identification risks. The DPA and product-specific terms should be reviewed for more precise AI-data restrictions.

2. Broader Commercial and Operational Changes

• Minimum commitments are more detailed and harder to reduce. They expressly cover agent seats, Add-Ons, phone numbers, telecommunications packages and usage-based AI-agent services. Minimum commitments remain billable even if actual usage is lower.
• Subscriptions generally cannot be downgraded or cancelled during the current term. Annual renewal applies unless at least three months’ non-renewal notice is provided.
• Fees increase automatically by 5% annually on a compounded basis, without separate notice. Dixa may also change list and overage prices, with overage pricing potentially effective immediately.
• Dixa may suspend access immediately for payment failures or breaches.
• Dixa may change the Services and Documentation and may amend the Terms with generally 30 days’ notice, although continued use constitutes acceptance.
• Data-loss protection is limited. The Customer’s sole remedy for lost or damaged data is commercially reasonable restoration from Dixa’s latest backup, and Dixa disclaims responsibility for third-party-caused loss.

The publisher records this document as revised on this date (“Last updated:28 May 2025”).

Summary of Important Changes

1. Privacy Policy title and website navigation expanded

• The heading was changed from a simple “Privacy Policy” and date reference to a broader Legal Hub navigation list, including:
• Terms of Service
• Cookie Policy
• GDPR
• Third-Party Services
• Data Processing Agreement
• Data Security
• Applicant Privacy Notice
• Global Feedback Panel
• Accessibility Statement
• Early Access Terms
• Referral Program
• The revised text expressly labels the document as “Privacy Policy” and adds “Last updated: 25 Novembre 2021.”

Risk/impact: This appears primarily to be website navigation and document-labeling content rather than a substantive privacy change. However, including links to other legal documents may make it easier for those documents to be treated as part of the overall contractual framework. Customers should review whether any linked terms contain broader data-use rights.

2. Incorporation into the Terms of Service clarified

• The language was revised from stating that the Privacy Policy is incorporated into and subject to the “Dixa Terms” to stating that it is incorporated into and subject to the “Dixa Terms of Service.”
• The document also states that it is subject to the Terms of Service and associated procedures, and that capitalized terms have the meanings given to them in the Policy.

Risk/impact: This reinforces contractual integration with the Terms of Service and potentially other procedures. Any conflicting or broader data-processing rights in the Terms of Service may apply. Customers should check priority-of-terms and amendment provisions.

3. Definitions largely retained, with minor drafting changes

• References to “Dixa,” “us,” “we,” the “Website,” “Dixa Services,” and “Customer Data” were reformatted and punctuation was standardized.
• “Personal Data” remains defined by reference to applicable legislation.

Risk/impact: No clear substantive expansion of the defined scope of Customer Data or Personal Data appears in the supplied diff. The formatting changes do not themselves create an evident new obligation.

4. Existing data-use disclosures remain

The diff retains provisions concerning use of information for:

• Analyzing usage trends and preferences
• Optimizing and administering the Services
• Monitoring traffic and user movements
• Gathering demographic information
• Corporate, shareholder, legal, and compliance purposes

5. AI-model training

• No express new language authorizing, restricting, or describing the use of Customer Data to train, fine-tune, evaluate, or develop AI models appears in this diff.
• The retained language about analyzing usage trends and optimizing Services is broad, but it does not expressly establish an AI-training right based on the supplied text.

Recommendation: Confirm the full current Privacy Policy, Terms of Service, and Data Processing Agreement for separate AI-training provisions, especially regarding customer content, de-identification, opt-out rights, retention, and use of data to improve models.

Summary of Important Changes and Risks

1. Major expansion and restructuring

The diff replaces a short version of the Terms of Service with a substantially expanded agreement containing detailed definitions and provisions covering:

• Services, accounts, permitted users and affiliates
• Add-ons and collaboration users
• Consultancy services
• Telephony, SMS and MMS
• Pricing, overage and minimum commitments
• Data protection and confidentiality
• Intellectual property
• Warranties, indemnities and liability
• Termination, renewals and governing law

Customers should review the full replacement terms rather than treating this as a limited wording update.

2. AI-related provisions and data use

New AI products and usage charges

“Add-On” now expressly includes AI products such as:

• Mim AI Co-Pilot
• AI Quality Assurance
• Advanced Insights

The definition of “Overage” expressly includes exceeding limits for data analyzed by AI tools or services. AI-related usage may therefore generate additional charges at then-current rates. The customer may be unable to reduce its minimum commitment during the subscription term, even if actual AI usage declines.

Use of aggregated and anonymized data

A new confidentiality provision allows Dixa to:

• Aggregate, collect and analyze data and information relating to the provision, use and performance of the Services, including traffic and telemetry data
• Use that information during and after the agreement to develop and improve the Services and other Dixa offerings
• Disclose it in aggregated and anonymized form
• Retain rights in the aggregated and anonymized data

The provision states that the data must not permit identification of an individual and that disclosures must not identify the customer or any individual.

AI model training risk

The terms do not expressly say that Dixa may use identifiable Customer Data, customer conversations, prompts, outputs or Personal Data to train general-purpose AI models. The apparent authorization is limited to data that has been aggregated and anonymized.

However, “develop and improve” the Services and other offerings is broad and could potentially include training, testing or tuning AI systems using derived or anonymized datasets. The agreement does not define:

• The anonymization standard
• Whether data will be used to train models shared across customers
• Whether AI inputs, outputs, transcripts or metadata are included
• Whether customers can opt out
• How re-identification risks are controlled
• How long derived datasets may be retained

Customers handling confidential, regulated or personal information should seek an express prohibition on model training or obtain specific contractual limits and an opt-out.

3. Data protection and confidentiality risks

The customer is expressly responsible for the legality, accuracy and integrity of Customer Data and must compensate Dixa for third-party royalties, fines and settlements arising from improperly supplied data.

Use of sensitive personal data is prohibited unless specifically agreed in the data processing agreement. Services may be operated outside the EU, subject to an approved transfer mechanism.

On termination, confidential information must generally be returned or destroyed, but the separate aggregated/anonymized data right survives and permits continued use after termination.

4. Commercial and operational risks

• Automatic 12-month renewals require at least three months’ non-renewal notice.
• A 5% annual fee increase applies automatically on a compounded basis.
• Dixa may change list and overage prices immediately through the Platform.
• Subscriptions generally cannot be downgraded or cancelled during the current term.
• Overage is non-cancellable and non-refundable.
• Dixa may suspend access for non-payment or breach.
• Liability is capped at fees paid during the preceding six months, with broad exclusions for data loss and consequential losses.
• Dixa may modify the Services and Documentation as it considers appropriate.
• Dixa receives rights to use the customer’s name and logo for marketing and may request a joint announcement.