accessiBe

Structured Summary of Changes

Overall change

The diff appears to replace a generic “Home Privacy Notice” presentation with an accessiBe-branded privacy notice interface. The substantive privacy-policy text appears unchanged across the versions shown. The same revised text is repeated in the supplied diff, so no additional substantive changes can be reliably identified.

Notable textual/interface changes

• Branding/navigation
• “Home” is replaced or supplemented with “accessiBe.”
• The header now includes “Your Data, Explained | accessiBe” and a “Skip content” accessibility/navigation element.
• The reference to the FAQ page appears to be replaced with a content/navigation link.
• These appear primarily to be website presentation and accessibility changes, rather than changes to privacy rights or data practices.

AI and model-training provisions

Existing AI-related disclosure

The notice states that accessiBe may:

• Record calls for customer support, quality assurance, training, and marketing.
• Process call recordings using AI tools to generate transcripts, summaries, and notes.
• Inform callers when recording begins and provide an opportunity to opt out of recording.

No express AI-training authorization

The notice does not expressly state that customer data, call recordings, chatbot inputs, website data, end-user data, or other personal information may be used to:

• Train, fine-tune, or improve generative AI or machine-learning models;
• Create general-purpose models;
• Share training data with AI vendors for their own model development; or
• Retain data for model-training purposes after the underlying service ends.

Accordingly, based on the supplied diff, there is no newly added permission to train AI models on customer data.

Residual AI-related risks

Even without an express training clause, the broad language allowing accessiBe to:

• “improve and optimize” its services;
• create aggregated, inferred, anonymized, or pseudonymized data; and
• use service providers to facilitate and enhance services

could potentially support limited machine-learning or analytics activities, depending on how those terms are interpreted and the terms applicable to vendors. The policy does not clearly define whether de-identified data may be used for AI training or whether AI vendors are prohibited from retaining or using submitted data for their own purposes.

Other material risks in the unchanged text

• Broad collection: Chatbot inputs, website structure, URLs, clicks, interactions, pages viewed, and accessibility preferences may be collected.
• Customer responsibility: Customers are responsible for notifying end users and obtaining consent for certain accessWidget processing.
• Broad retention: Data may be retained beyond account closure for legal, audit, dispute, and enforcement purposes.
• International transfers: Data may be transferred to the United States and other jurisdictions.
• Marketing: One-time-service users may receive marketing communications for up to six months, based on a contractual-necessity theory that could be challenged in some jurisdictions.
• Policy changes: Substantial changes may become accepted after ten days’ notice unless the user stops using the Services.

Summary of Important Changes

Overall assessment

The diff does not appear to contain substantive changes to the Privacy Notice’s privacy, data-use, disclosure, retention, or AI provisions. The displayed text is otherwise materially identical across the versions, and the repeated blocks appear to reflect formatting or comparison artifacts rather than additional legal amendments.

Identified changes

Branding and navigation

• “Home” is replaced with “accessiBe.”
• This appears to be a branding or navigation-label change.
• It does not change the parties covered by the notice, which remain accessiBe Ltd. and its affiliates, including accessiBe Inc.

• Navigation wording is changed:
• “our” is replaced with “content.”
• These appear to be website interface or accessibility-navigation changes rather than changes to legal rights or obligations.

AI model training and AI-related data use

• The notice continues to state that calls may be:
• recorded for customer support, quality assurance, training, and marketing; and
• processed using AI tools to generate transcripts, summaries, and notes.

• No new language was identified authorizing the use of customer data, call recordings, transcripts, website data, end-user data, or other Personal Information to train AI models.
• No deletion or limitation of an existing AI-training restriction was identified.
• The notice does not expressly address:
• whether data may be used to train general-purpose or third-party AI models;
• whether AI vendors may retain or reuse submitted data;
• whether customer data is used for model fine-tuning;
• whether customers can opt out of AI processing; or
• whether AI-generated outputs are used for profiling or automated decision-making.

Existing risks that remain

Although not introduced by this diff, the notice continues to present several points for review:

• Broad service-provider access: Service Providers may access Personal Information for their stated roles, but the notice does not identify AI providers or specify whether their contractual terms prohibit model training or secondary use.
• Call-recording exposure: Calls may be recorded and processed by AI tools. The notice provides an opt-out opportunity, but does not explain how opting out affects service availability or how long recordings and AI-generated materials are retained.
• Customer responsibility for end-user data: Customers are responsible for notifying end users and obtaining consent for certain accessWidget processing.
• Broad improvement language: Information may be used to improve and optimize Services and create aggregated, anonymized, or pseudonymized data. The boundary between service improvement and AI development is not clearly defined.

Bottom line

The apparent changes are limited to branding and navigation text. No material legal change or change concerning AI-model training is discernible from the supplied diff. Nevertheless, the notice would benefit from expressly stating whether any customer or end-user data may be used to train, fine-tune, evaluate, or improve AI models, including models operated by third-party providers.

Summary of Changes

Overall change

• Approximately 5,967 words were removed from the document.
• No replacement or newly added language is shown.
• Because the deleted text is not provided, the specific legal provisions affected cannot be identified with certainty.

Key legal risks

Loss of protections

The deleted material may have contained important customer protections, including:

• Confidentiality and data-security obligations
• Limits on the provider’s use of customer data
• Data ownership and intellectual-property rights
• Restrictions on disclosure to third parties
• Data retention and deletion requirements
• Indemnities, warranties, or liability limits
• Audit, compliance, or breach-notification rights
• Termination and post-termination obligations

Removing these provisions could materially shift risk to the customer, particularly if no replacement language was added elsewhere.

Interpretation risk

Because the diff identifies only a bulk deletion, it is not possible to determine whether:

• The deleted terms were redundant or replaced in another section;
• The document now contains gaps or conflicting provisions;
• Previously negotiated customer rights have been removed; or
• The deletion changes the commercial balance of the agreement.

The complete prior and revised versions should be compared before accepting the change.

Customer data and AI-model training

• The supplied diff does not identify whether any language about using customer data to train AI models was added, removed, or modified.
• However, if the deleted 5,967 words included restrictions on artificial-intelligence use, the deletion could potentially allow broader use of customer data, prompts, outputs, telemetry, or metadata for:
• Training or fine-tuning models;
• Improving services;
• Developing products;
• Human review or safety testing; or
• Sharing with affiliates or third-party service providers.
• Conversely, the deletion may have removed an express prohibition or opt-out right, making the provider’s data-use rights less clear and potentially increasing customer exposure.

Recommended review points

The revised agreement should be checked specifically for clear provisions stating:

1. Whether customer data may be used to train, fine-tune, or improve AI models;

2. Whether training is enabled by default or requires consent;

3. Whether the customer can opt out;

4. Whether inputs, outputs, personal data, and usage metadata are treated differently;

5. Whether data is anonymized or de-identified;

6. Whether data is shared with affiliates or subprocessors; and

7. How data is retained and deleted after termination.

Bottom line: The only confirmed change is a substantial deletion. Its legal effect—and any impact on AI training—cannot be reliably assessed without the deleted text or the complete before-and-after agreements.

Summary of Changes

Content Removed

• The phrase “Connecting you with the right rep…” was removed.
• This appears to be a navigation, marketing, or user-interface phrase rather than a substantive privacy or contractual provision.

Data Use and AI Training

• The provided diff contains no changes addressing:
• Whether customer data may be used to train AI or machine-learning models;
• Whether prompts, inputs, outputs, or customer content may be retained;
• Whether data is used to improve accessiBe’s products or services;
• Whether data is shared with AI providers or other third parties;
• Opt-out rights or restrictions on AI-related processing; or
• De-identification, aggregation, or anonymization for model training.

Legal and Commercial Impact

• Based solely on the text provided, there is no identifiable change to privacy rights, data-processing permissions, confidentiality obligations, or AI-training rights.
• The change may affect website presentation or navigation, but it does not appear to create a new legal obligation or customer risk.

Important Limitation

• The diff appears incomplete and contains only a page title and one deleted phrase. A reliable review of changes to customer-data use—particularly AI-model training—would require the portions containing the privacy notice’s data-use, service-provider, retention, and machine-learning provisions.

Summary of Changes

Deleted Content

• The phrase “Connecting you with the right rep...” was removed.
• Based on the limited diff provided, this appears to be website navigation or marketing text rather than a substantive contractual term.

Added Content

• No substantive additions are shown in the diff.

AI Training and Customer Data

• The provided diff contains no language addressing:
• Use of customer data to train, fine-tune, or improve AI models;
• Use of customer content for machine learning or generative AI;
• Whether customer data is used in aggregated, de-identified, or identifiable form;
• Customer rights to opt out of AI training;
• Ownership or licensing of AI-generated outputs; or
• Security, retention, or deletion of data used for AI purposes.

Accordingly, no change regarding AI model training or customer-data usage can be identified from the supplied excerpt.

Legal and Commercial Risk

• The excerpt does not show any change to contractual rights, obligations, warranties, indemnities, data-processing terms, intellectual-property provisions, or liability allocation.
• The deleted phrase may affect website presentation or lead-generation functionality, but it does not appear to create a material legal risk on its face.
• This conclusion is limited because the diff appears incomplete. A full comparison of the Terms of Service would be necessary to identify changes to data use, AI training, privacy, or other legal provisions.

Summary

• The diff only states that approximately six words were removed.
• The specific deleted language is not provided, so the legal effect cannot be determined.
• No additions, replacements, or identifiable changes concerning:
• Customer data ownership or rights
• Permitted data uses
• Confidentiality or privacy
• AI or machine-learning model training
• Retention, deletion, or security obligations
• Customer consent or opt-out rights

AI Training Impact

There is not enough information to conclude that the document’s treatment of customer data for AI training has changed. The six deleted words could theoretically affect such provisions, but this cannot be assessed without seeing the actual text removed and the surrounding clause.

Risk Assessment

• Current assessment: No identifiable new legal risk based on the information supplied.
• Key limitation: An isolated deletion can materially change meaning—for example, by removing a qualifier such as “only,” “with consent,” “aggregated,” or “not for training.”
• Recommended next step: Provide the actual deleted wording and the surrounding provision, particularly any clause addressing data use, analytics, artificial intelligence, machine learning, or service improvement.

Summary

The diff only states that approximately six words were removed from the document. It does not identify:

• Which words were deleted;
• The clause or section affected;
• Whether the deletion changes the parties’ rights or obligations; or
• Whether any provisions concerning customer data, artificial intelligence, machine learning, or model training were modified.

AI Training and Data Use

No specific change related to the use of customer data to train AI models can be identified from the information provided. In particular, the diff does not show whether the document now:

• Permits or restricts training AI models on customer data;
• Allows the provider to use customer content, prompts, outputs, metadata, or usage information for training;
• Requires data to be anonymized or de-identified before training;
• Gives the customer an opt-out right;
• Allows human review or sharing with affiliates or third-party model providers; or
• Creates ownership, confidentiality, security, or deletion risks relating to training data.

Risk Assessment

Because the deleted language is not shown, the legal significance cannot be determined. Even a short deletion could materially affect:

• Consent or authorization to use data;
• Confidentiality protections;
• Restrictions on secondary use;
• Data retention or deletion obligations;
• Allocation of intellectual-property rights; or
• Liability for unauthorized use.

The redline should be reviewed with the actual deleted words and surrounding clause. On the current information, no reliable conclusion can be reached about whether the change increases or decreases the customer’s exposure to AI-training or other data-use risks.

Summary

The diff only states:

> “Added approximately 5967 words to the document”

No actual contractual language, additions, deletions, or replacements are provided. Accordingly, the substantive legal impact cannot be assessed.

Key Findings

• No identifiable changes: The diff does not show which provisions were added, removed, or modified.
• No AI-training provisions visible: There is no language addressing whether customer data may be:
• Used to train, fine-tune, or improve AI models;
• Shared with AI service providers or subprocessors;
• Used in aggregated, anonymized, or de-identified form;
• Retained after termination;
• Excluded from model training by default or only through opt-out;
• Used to generate outputs that may benefit other customers.
• No risk assessment possible: It is not possible to determine whether the new language creates risks involving confidentiality, ownership, privacy, security, data retention, regulatory compliance, or intellectual property.
• Potential process concern: A statement that approximately 5,967 words were added, without the text itself, prevents verification of the scope and effect of the changes. The added material could include significant new obligations or rights.

Information Needed

Please provide the actual redlined text using the stated notation:

• Additions: {text}
• Deletions: [text]
• Replacements: [old text]{new text}

The full diff is necessary to identify changes concerning customer-data use, AI-model training, data ownership, confidentiality, subprocessors, retention, and related legal risks.

Summary of Important Changes

Overall assessment

The diff appears to contain repeated or duplicated versions of substantially the same Privacy Notice. No clear substantive change to the main data-use provisions is identifiable beyond website/navigation wording and the inclusion of AI-related call-processing language.

AI and model-training provisions

What the notice expressly says

The Privacy Notice states that accessiBe may:

• Record telephone calls for customer support, quality assurance, training, and marketing.
• Process call recordings using artificial intelligence tools to generate:
• Transcripts;
• Summaries; and
• Notes.
• Inform callers when recording begins and provide an opportunity to opt out of recording.

What it does not say

The notice does not expressly state that customer data, call recordings, transcripts, chatbot inputs, website data, or end-user data will be used to:

• Train general-purpose AI models;
• Fine-tune proprietary or third-party AI models;
• Improve AI models for other customers; or
• Permit AI vendors to retain data for their own model-training purposes.

Accordingly, this diff does not appear to add an explicit customer-data-to-AI-training right. However, the wording is not fully protective because:

• “Training” is listed as a purpose for call recording and could be interpreted as employee or service training, but it is not defined.
• “Improve and optimize the Services” and use of data to create aggregated, inferred, anonymized, or pseudonymized data are broad enough to support certain AI-development activities, particularly if data is claimed to be anonymized.
• “Service Providers” may access personal information for their stated roles, but the notice does not identify AI providers or expressly prohibit those providers from using data to train their own models.
• The notice says accessiBe processes all information entered into its chatbot, which may include confidential customer information, but provides no AI-specific restriction or retention detail for chatbot data.
• Website and accessWidget data—including HTML structure, URLs, clicks, interactions, and pages viewed—may be used to improve services. The notice does not clarify whether these data may be used in machine-learning systems.

Other notable risks

• Broad consent by use: In jurisdictions requiring consent, browsing or using the Services is deemed consent to all purposes in the notice.
• Customer responsibility: Customers are responsible for informing end users and obtaining consent for processing accessWidget data.
• Broad retention: Data may be retained beyond account closure for legal, audit, dispute, enforcement, or business purposes, with no precise schedule for many categories.
• Broad sharing: Personal information may be shared with affiliates, service providers, and transaction participants in a merger or acquisition.
• Marketing arrangement: Users receiving one-time services may be treated as contractually agreeing to marketing for up to six months.

Recommended clarification

Customers should seek written confirmation that their data, end-user data, chatbot submissions, recordings, transcripts, and derived data will not be used to train or improve general-purpose AI models—or disclosed to AI providers for those purposes—unless expressly authorized.

Executive Summary

The diff appears to make website/navigation and branding changes only. The substantive privacy notice language—including provisions concerning AI processing, data use, sharing, retention, and customer/end-user data—appears unchanged.

Important Changes

1. Branding and page navigation

• “Home” is replaced with “accessiBe.”
• The prior effective-date/FAQ navigation text is replaced with:
• “Your Data, Explained”
• “accessiBe Skip”
• A link reference changes from “our FAQ page” to “content.”

These changes appear editorial or user-interface related and do not alter the stated legal rights or obligations.

AI and Model-Training Provisions

No new AI-training authorization identified

The diff does not add language expressly permitting accessiBe to use customer data, end-user data, website content, call recordings, transcripts, or other Personal Information to:

• train, fine-tune, or improve artificial-intelligence models;
• develop generalized machine-learning systems;
• create datasets for AI training; or
• permit third-party AI providers to use such data for their own model training.

Existing AI processing remains

The notice continues to state that accessiBe may:

• record calls for customer support, quality assurance, training, and marketing;
• process call recordings using AI tools to generate transcripts, summaries, and notes; and
• inform callers and provide an opportunity to opt out of recording.

This is an authorization for AI-assisted call processing, but it is not expressly an authorization for AI model training. However, the wording does not clearly state whether AI vendors may retain or use recordings, transcripts, or derived information to train their own models. That ambiguity remains a potential vendor-management and transparency risk.

Other Continuing Risks

• accessWidget-related data—including URLs, HTML structure, CSS attributes, clicks, interactions, pages viewed, IP addresses, and accessibility preferences—continues to be collected or processed.
• accessiBe continues to characterize itself as a processor for customer user data, while customers remain responsible for disclosures and obtaining any required end-user consent.
• Broad purposes remain permitted, including improving services, creating aggregated or anonymized data, marketing, analytics, and sharing with service providers and affiliates.
• Cross-border transfers and third-party service-provider access remain broadly described.

Bottom Line

Based on the supplied diff, there is no material change to AI-training rights. The principal changes are branding and navigation updates. The existing AI-related call-processing language remains, with no express model-training permission but also no explicit contractual restriction on AI providers’ retention or secondary use.

Structured Summary of Important Changes

1. Document and presentation changes

• The notice title changes from “accessiBe | Your Data, Explained | accessiBe Skip” to “Home Privacy Notice.”
• The navigation and introductory wording are revised, including a link to the company’s FAQ page and a change from “go to content” to “go to our FAQ page.”
• The revised notice is expressly dated June 28, 2026 and states that it is incorporated into the Terms of Service and other agreements.

These appear primarily editorial, but incorporation by reference makes the notice contractually significant.

2. AI-related processing

The revised text expressly states that accessiBe may:

• Record calls for customer support, quality assurance, training, and marketing.
• Process call recordings using artificial intelligence tools to generate transcripts, summaries, and notes.
• Inform callers at the beginning of the call and provide an opportunity to opt out of recording.

Key AI risk

The notice does not expressly authorize or describe using customer data, call recordings, transcripts, website content, end-user data, or support-chat information to train, fine-tune, or improve generative AI or other machine-learning models. It also does not state:

• Whether AI providers may retain or use the data for their own model training;
• Whether data is de-identified before AI processing;
• Which AI vendors receive the information;
• How long AI inputs and outputs are retained;
• Whether customers can opt out of AI processing separately from call recording; or
• Whether customer website content or end-user information is used to improve accessiBe’s models.

Accordingly, the notice appears to permit AI-assisted transcription and summarization, but does not clearly grant a separate model-training right. The broad language allowing accessiBe to use information to “improve and optimize the Services,” create inferences, and share information with service providers could nevertheless create ambiguity and should be clarified contractually.

3. Expanded data collection and use

The notice describes collection of substantial data, including:

• Website HTML structure, CSS attributes, URLs, clicks, interactions, and pages viewed through accessWidget;
• End-user IP addresses, URLs, and accessibility preferences;
• Chatbot and customer-support communications;
• Analytics, cookie, advertising, and social-media information; and
• Information obtained from marketing providers and other third parties.

AccessiBe may combine information from multiple sources and create aggregated, inferred, anonymized, or pseudonymized data for its own and business partners’ use.

4. Customer compliance obligations

Customers are made responsible for using commercially reasonable efforts to:

• Disclose accessWidget data processing clearly to end users; and
• Obtain end-user consent where required.

AccessiBe acts as a processor for end-user data under a separate Data Processing Addendum. This may shift notice, consent, and regulatory-compliance risk to the customer.

5. Other material risks

• Data may be transferred and stored in the United States, EU, UK, Australia, and additional locations selected by accessiBe.
• Data may be shared with service providers, affiliates, law-enforcement authorities, and transaction counterparties.
• One-time-service users may receive marketing for up to six months as part of the contractual exchange for the service.
• Retention periods are flexible and may extend beyond account closure for legal, audit, dispute, and enforcement purposes.
• California users’ data may be “shared” with third-party cookie providers, although accessiBe states it does not “sell” personal information.

Summary

The provided diff contains only the statement:

> “Added approximately 5967 words to the document”

It does not include the actual added language, deleted language, or replacement text. As a result, the legal and commercial impact cannot be reliably assessed.

AI Training and Data Use

• No clause text is provided addressing whether customer data may be:
• Used to train, fine-tune, or improve AI models;
• Used to develop products or services;
• Combined with other customers’ data;
• Reviewed by humans or shared with service providers;
• Retained after termination;
• Anonymized, aggregated, or otherwise de-identified;
• Used for model testing, monitoring, or safety purposes.
• It is therefore impossible to determine whether the changes expand or restrict the provider’s rights to use customer data.
• The missing language could materially affect confidentiality, ownership, privacy, security, regulatory compliance, and the customer’s ability to opt out of AI training.

Other Legal Risks That Cannot Yet Be Assessed

The missing additions may also change:

• Intellectual-property ownership and licensing;
• Confidentiality obligations and permitted disclosures;
• Data-processing and privacy commitments;
• Security standards and breach-notification obligations;
• Data retention and deletion rights;
• Warranties, indemnities, and liability caps;
• Service levels, suspension, and termination rights;
• Audit, compliance, and subcontracting rights;
• Governing law and dispute-resolution procedures.

Required for Further Review

Please provide the actual redlined text or the full revised clauses. In particular, include any provisions containing terms such as “customer data,” “content,” “input,” “output,” “train,” “fine-tune,” “improve,” “machine learning,” “artificial intelligence,” “model,” “aggregate,” “de-identify,” or “service improvement.”

Summary of Changes

Scope of the Diff

• The diff indicates that approximately 5,967 words were removed from the document.
• No replacement language or the text of the deleted provisions was provided.
• Because the deleted text is unavailable, the specific legal and commercial effects cannot be determined reliably.

Key Risks

• Loss of contractual protections: The removed provisions may have addressed confidentiality, security, service levels, warranties, liability, indemnification, audit rights, data ownership, or termination rights.
• Potentially broader discretion for the provider: If limitations or restrictions were deleted without replacement, the remaining agreement may give the provider greater flexibility in handling customer data or changing services.
• Unclear allocation of risk: Deleted provisions may have limited liability or established remedies. Their removal could either increase the customer’s exposure or, depending on the surrounding language, create ambiguity that is difficult to enforce.
• Inconsistent cross-references: Removing a large block of text can leave references to deleted sections, undefined terms, or conflicting obligations elsewhere in the agreement.

Customer Data and AI Training

• No AI-training change can be confirmed from the information provided.
• The diff does not identify whether the deleted provisions:
• prohibited or permitted using customer data to train AI models;
• allowed use of prompts, inputs, outputs, telemetry, or metadata for training;
• required de-identification, aggregation, or anonymization;
• restricted human review or provider access;
• granted the customer an opt-out right; or
• addressed ownership of models, model outputs, or derived data.
• However, deleting data-use restrictions, confidentiality language, or purpose limitations could create a material risk that customer data may be used more broadly than before—particularly if the surviving agreement contains general rights to use data for service improvement, analytics, or product development.

Recommended Review

Before accepting the revised document, compare the prior and current full versions and specifically verify:

1. Whether customer data may be used to train, fine-tune, evaluate, or improve AI models.

2. Whether the customer has an express opt-out or prohibition right.

3. Whether prompts, outputs, personal data, and metadata are covered.

4. Whether data must be de-identified and deleted after use.

5. Whether subcontractors or third-party AI providers may access the data.

6. Whether ownership, confidentiality, security, and liability protections remain intact.

A reliable legal assessment requires the actual deleted language and the provisions that remain.

Executive Summary

The diff appears to contain substantial duplicated text, but there are very few identifiable substantive changes. The principal apparent changes are branding/navigation updates and the addition of an FAQ reference. No express authorization to use Customer Data to train AI models is visible.

Identifiable Changes

1. Branding and page-navigation changes

• “accessiBe” appears to be changed to “Home” in the page header.
• The prior title/navigation text “| Your Data, Explained | accessiBe Skip” appears to be replaced with “Privacy Notice”.
• The notice now refers users to “our FAQ page”, replacing or supplementing the prior generic “content” link.

Risk: These appear primarily editorial or website-navigation changes. However, the FAQ may contain additional privacy or AI-related disclosures. Customers should review the linked FAQ because incorporation or reliance on external website content could create ambiguity about which terms apply.

2. Effective date

• The notice states “Effective as of: June 28, 2026.”

Risk: This date should be confirmed as intentional. If the notice is being issued before that date, or if it is intended to replace an earlier notice immediately, there may be uncertainty about when the updated processing terms become effective.

AI and Model-Training Analysis

No express AI-training authorization identified

The diff does not appear to add language expressly permitting accessiBe to:

• use Customer Data to train, fine-tune, or improve general-purpose AI models;
• use Customer Data for model development beyond providing the Services;
• retain prompts, outputs, transcripts, or other inputs for AI training;
• share Customer Data with AI providers for their own model-training purposes; or
• use one customer’s data to train models benefiting other customers.

Accordingly, there is no clear new contractual grant of AI-training rights in the visible changes.

Existing AI-related processing remains important

The notice states that recorded calls may be processed with AI tools to create transcripts, summaries, and notes. This is operational AI processing, but it is not expressly described as model training.

Potential risks remain because the notice does not clearly state:

• which AI providers process recordings;
• whether providers may retain or use data for their own training;
• whether call data includes confidential Customer information;
• how long transcripts and AI-generated notes are retained; or
• whether Customers can prohibit AI processing other than opting out of call recording.

The notice also permits broad use of information to improve and optimize the Services and to create aggregated, anonymized, or pseudonymized data. Depending on the Terms, this language could potentially support machine-learning or model-improvement activities, although it does not expressly say so.

Recommended Review Points

• Confirm whether the Terms, FAQ, DPA, or AI-provider terms contain separate model-training permissions.
• Require express limits on AI-provider retention and secondary use.
• Clarify whether Customer Data, End-User Data, call recordings, transcripts, and outputs may be used for training.
• Confirm that anonymization and deletion standards are sufficiently protective and verifiable.

Summary

Key limitation

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

> “Added approximately 5967 words to the document”

Without the text of the additions, deletions, or replacements, it is not possible to reliably determine:

• What contractual obligations have changed;
• Whether liability, indemnity, warranties, termination, confidentiality, or security provisions were modified;
• Whether customer data may be used to train, fine-tune, evaluate, or improve AI models;
• Whether data may be shared with affiliates, vendors, or other third parties;
• Whether the customer can opt out of AI training or request deletion;
• Whether any new intellectual-property rights are granted to the provider; or
• Whether the provider has expanded rights to retain or create derivative data.

AI-training risks to check

The full diff should be reviewed specifically for language allowing the provider to:

1. Use customer data for model training or improvement

Look for terms such as “train,” “fine-tune,” “improve,” “develop,” “optimize,” “machine learning,” “AI models,” or “services improvement.”

2. Use data after termination

Check whether training rights continue after the agreement ends or after data is deleted.

3. Use de-identified, aggregated, or derived data

Determine whether the provider can use supposedly anonymized data and whether re-identification is prohibited.

4. Share data with third parties

Review rights involving subprocessors, affiliates, cloud providers, model providers, or public AI systems.

5. Claim ownership or broad licenses

Identify any license allowing the provider to copy, modify, create derivative works from, commercialize, or distribute customer data or outputs.

6. Limit customer protections

Check for disclaimers concerning confidentiality, security, accuracy, regulatory compliance, or responsibility for model outputs.

Information needed

Please provide the actual redlined text, including all additions, deletions, and replacements. The current submission contains no substantive contractual language to analyze.

Summary

The provided diff is incomplete. It contains only the following text:

> “accessiBe Terms of Service | Legal Agreement | accessiBe {Skip to content} Connecting you with the right rep...”

No substantive contractual provisions are included, so it is not possible to reliably identify legal changes or assess how customer data may be used to train AI models.

Changes Identifiable from the Provided Text

• “Skip to content” added
• This appears to be a website accessibility/navigation feature rather than a contractual term.
• It does not appear to change the parties’ rights, obligations, data-use permissions, liability, or other legal terms.

AI Training and Customer Data

• No AI-training language is present in the supplied excerpt.
• The excerpt does not show whether accessiBe may:
• Use customer data to train, fine-tune, or improve artificial intelligence or machine-learning models;
• Share customer data with AI providers or subprocessors;
• Use aggregated, de-identified, or anonymized data for model development;
• Retain customer data after termination for AI-related purposes; or
• Allow customers to opt out of such uses.

Risk Assessment

No new contractual risks can be assessed from the excerpt. In particular, the relevant provisions would likely be found under headings such as:

• Data Processing or Privacy
• Customer Data
• Confidentiality
• Service Improvement
• Artificial Intelligence or Machine Learning
• Subprocessors
• License or Usage Rights
• Retention and Deletion
• Termination

A complete before-and-after diff, including the surrounding provisions and the deleted text, is needed for a meaningful legal analysis.

Summary

The provided diff is extremely limited. It shows only the following apparent addition:

• Added: “Skip to content”

Legal and Privacy Impact

• No substantive privacy or contractual changes are visible.
• The addition appears to be an accessibility/navigation feature that allows users to bypass repeated page elements and move directly to the main content.
• This change does not, on its face, alter:
• What personal data is collected;
• How data is used, disclosed, or retained;
• Customer rights or obligations;
• Security commitments;
• Cookies or tracking technologies; or
• Liability, indemnification, or other legal terms.

AI Training and Model Use

• No language concerning AI models, machine learning, model training, data processing for training, or use of customer data to improve AI systems appears in the supplied diff.
• Accordingly, the diff does not show any new authorization for accessiBe to use customer data to train AI models.
• However, because the provided text appears incomplete—ending with “Connecting you with the right rep...” —no conclusion can be drawn about changes elsewhere in the Privacy Notice.

Overall Assessment

Based solely on the text provided, the change is non-substantive and accessibility-related. A complete before-and-after diff is needed to assess privacy, data-use, or AI-training risks.

Summary

The provided diff only states that approximately 11 words were removed, without identifying the deleted language or showing any replacement text.

Legal and Commercial Impact

• Cannot determine the substantive effect of the deletion.
• The change could affect important terms such as:
• Customer data ownership or usage rights
• Confidentiality obligations
• Data retention or deletion
• Security commitments
• Liability, indemnification, or warranties
• Termination rights
• Governing law or dispute resolution

AI Training and Data Use

• No conclusion can be reached regarding AI model training.
• The diff does not show whether language was added, removed, or changed concerning:
• Using customer data to train, fine-tune, or improve AI models
• Using customer prompts, inputs, outputs, or metadata for model development
• Opt-out or consent requirements
• Aggregation, anonymization, or de-identification
• Human review or provider access to customer data
• Restrictions on using confidential or personal information

Risk Assessment

The lack of the actual 11 deleted words creates a material review limitation. If the deleted language narrowed restrictions on data use, removed an AI-training prohibition, or expanded the provider’s rights, the change could increase the customer’s privacy, confidentiality, intellectual-property, and regulatory risks.

The complete before-and-after wording is required for a reliable legal analysis.

Summary of Changes

Deleted Website Script

The diff removes a JavaScript event listener from the “Your Data, Explained” page:

```javascript

document.addEventListener('click', (e) => {

if (e.target.closest('[ ]')) {

e.preventDefault();

e.stopPropagation();

}

}, true);

```

This code appears to intercept clicks on elements matching the selector "[ ]" and prevent the associated action from occurring. Because the selector is incomplete or malformed as shown, its exact operational effect cannot be determined.

Potential Legal and Practical Impact

• User-interface behavior: Removing the script may allow previously blocked links, buttons, or other interactive elements to function normally.
• Consent or disclosure controls: If the script was intended to prevent access to, or interaction with, a consent mechanism, privacy notice, opt-out control, or data settings interface, its removal could change how users provide or manage consent.
• Compliance uncertainty: The diff does not show whether any privacy terms, data-processing provisions, or contractual rights have changed. The business should verify that any newly enabled controls accurately reflect the governing privacy policy and applicable consent requirements.
• No apparent substantive contract change: The provided diff does not add or remove legal language concerning data collection, sharing, retention, security, or customer rights.

AI Training and Model Use

No change relating to the use of customer data to train artificial-intelligence or machine-learning models is visible in this diff.

Specifically, the diff does not show any new or removed language concerning:

• Training, fine-tuning, or improving AI models
• Use of customer content, inputs, outputs, or personal data for model development
• Opt-out or objection rights regarding AI training
• De-identification or aggregation for AI purposes
• Sharing data with AI providers or subprocessors

Because the diff is limited and appears to contain only a deleted script, it is not possible to determine whether the surrounding privacy or AI-training terms remain unchanged elsewhere in the document.

Summary

The diff only states:

> “Added approximately 11 words to the document”

It does not identify the actual words added, nor does it show any deletions or replacements. As a result, the legal effect of the amendment cannot be reliably analyzed.

AI Training and Customer Data

• No specific language is provided concerning:
• Whether customer data may be used to train, fine-tune, or improve AI models;
• Whether prompts, inputs, outputs, personal information, or confidential information may be retained;
• Whether data may be shared with affiliates, vendors, or third-party model providers;
• Whether customer data may be used in aggregated, anonymized, or de-identified form;
• Whether customers can opt out of AI training or require deletion; or
• Ownership, licensing, confidentiality, or security protections for data used in AI systems.

Accordingly, no conclusion can be drawn about whether the amendment expands or restricts the provider’s rights to use customer data for AI training.

Risk Assessment

The principal risk is lack of visibility into the actual amendment. Even a short addition could materially change:

• The provider’s license to use customer content;
• The purposes for which data may be processed;
• The duration of data retention;
• Disclosure or transfer rights;
• Confidentiality obligations;
• Allocation of intellectual-property rights; or
• Compliance obligations relating to privacy and regulated data.

Recommended Next Step

Obtain the actual 11-word addition, together with the surrounding clause and any associated deletions or replacements. The amendment should then be reviewed specifically for language such as “train,” “improve,” “develop,” “machine learning,” “artificial intelligence,” “models,” “inputs,” “outputs,” “content,” “aggregated,” “de-identified,” or “service improvement.”

Summary

The supplied diff only states: “Added approximately 11 words to the document.” It does not include the actual added language or identify where the change appears.

Important Changes

• No substantive contractual change can be analyzed from the information provided.
• The identity, location, and legal effect of the 11 added words are unknown.
• It is therefore not possible to assess whether the addition changes:
• Customer obligations or rights
• Data ownership or licensing
• Confidentiality or security requirements
• Liability, indemnification, or termination rights
• Fees, renewals, or service scope
• Governing law or dispute procedures

AI Training and Data Use

• The diff does not reveal whether customer data may be used to train, fine-tune, evaluate, or improve artificial intelligence or machine-learning models.
• No conclusion can be reached about whether:
• Customer data may be used for provider-wide model training
• Use is limited to providing the services
• Customer consent or an opt-out is required
• De-identified or aggregated data may be used
• Inputs, outputs, prompts, or usage metadata are retained
• The provider may share data with AI subcontractors or third parties

Risk Assessment

Unable to assess. The actual 11-word addition is required to identify new legal risks. The complete redlined text, including the words shown in braces, brackets, or replacement notation, should be provided for review.

Summary of Available Change

The diff states only that approximately 5,964 words were removed from the document. It does not identify which clauses, definitions, rights, or obligations were deleted.

Key Legal Implications

• Material uncertainty: A deletion of this size could remove important provisions concerning liability, confidentiality, data protection, security, intellectual property, warranties, termination, governing law, or dispute resolution.
• Potential loss of customer protections: If customer-favorable terms were removed, the customer may have reduced contractual protections without the specific changes being apparent.
• Potential loss of provider rights or obligations: The deletion could also remove restrictions on the provider, including data-use limits, security commitments, audit rights, or service-level obligations.
• Interpretation risk: Removing definitions or cross-referenced provisions may change the meaning of provisions that remain in the agreement or create internal inconsistencies.

AI Training and Data Use

The supplied diff does not show the deleted language. Therefore, it is not possible to determine whether the changes:

• Permit or restrict using customer data to train, fine-tune, or improve AI models;
• Permit use of prompts, inputs, outputs, usage data, or personal information for model development;
• Require customer consent or provide an opt-out;
• Limit training to aggregated, anonymized, or de-identified data;
• Give the provider ownership or broad license rights over customer data or outputs; or
• Impose deletion, retention, confidentiality, or security obligations relating to AI training data.

Because approximately 5,964 words were removed, any deleted AI or data-use provisions should be treated as potentially significant until reviewed.

Recommended Review

Provide either:

1. The full redline showing the deleted text; or

2. The original and revised versions of the agreement.

The deletion should be specifically checked for clauses containing terms such as “train,” “training,” “improve,” “develop,” “machine learning,” “AI,” “models,” “inputs,” “outputs,” “customer data,” “usage data,” “aggregate,” “de-identify,” “license,” “consent,” and “opt out.”

Summary

The provided diff states only: “Removed approximately 3 words from the document.”

Key Findings

• The specific deleted words are not identified.
• No additions, replacements, or substantive clause text are shown.
• It is therefore impossible to determine whether the change affects:
• Customer data rights or ownership
• Permission to use customer data
• Confidentiality or security obligations
• Data retention or deletion
• Liability, indemnification, or compliance obligations
• Use of customer data to train, fine-tune, evaluate, or improve AI models

AI Training/Data-Use Risk

No conclusion can be reached about changes to AI-model training. The deleted language could potentially narrow or expand the provider’s rights to use customer data, but the actual impact cannot be assessed without the missing text.

Recommended Follow-Up

Provide the exact three deleted words and the surrounding sentence or clause, preferably using the original diff notation. This is necessary to determine whether the deletion changes the customer’s consent, the provider’s data-use rights, or any restrictions on AI training.

Summary of Important Changes

1. AI-related processing of customer data

Call recording and AI processing added

The notice now states that accessiBe may:

• Record calls for customer support, quality assurance, training, and marketing.
• Process recordings using AI tools to create transcripts, summaries, and notes.
• Inform callers at the start of the call when recording occurs.
• Provide an opportunity to opt out of recording.

Risks and implications:

• Call content may contain confidential business information, personal data, or sensitive information supplied by the customer.
• The notice does not clearly identify the AI providers, whether recordings or transcripts are transferred to third parties, how long they are retained, or whether they are used to train general-purpose AI models.
• It does not expressly state that customer call data will not be used to train AI models. The reference to “training” appears to concern employee or service-quality training, but the language could be interpreted broadly.
• Customers may need to provide their own call-recording notices and obtain legally required consent from participants, particularly in jurisdictions requiring all-party consent.

No express authorization to train models identified

The diff does not appear to add an express provision authorizing accessiBe to use customer data, website content, end-user data, support communications, or call recordings to train AI models.

However, the new AI-processing language is incomplete from a risk perspective because it does not expressly distinguish between:

1. Using AI tools to process data and generate outputs; and

2. Retaining or using that data to train, improve, or fine-tune AI models.

Customers should seek confirmation that their data will not be used for model training unless separately authorized.

2. End-user data and customer responsibility

The notice states that accessiBe may process end-user IP addresses, URLs, accessibility preferences, and other information while users operate accessWidget. The customer is responsible for making commercially reasonable efforts to disclose this processing and obtain consent.

AccessiBe characterizes itself as a processor of customer user data, with processing governed by the Data Processing Addendum.

Risk: The customer bears an explicit compliance burden for notices and consent involving its website users. The scope of “other information that is not Personal Information” is also broad and may require further clarification.

3. Marketing use of data from one-time services

For limited or one-time services, such as accessScan, the customer agrees that accessiBe may use submitted contact information for marketing communications for up to six months.

The notice characterizes this as contractual consideration under GDPR Article 6(1)(b).

Risk: Treating marketing permission as part of the service contract may be challenged in some jurisdictions, where consent or a separate legitimate-interest analysis may be required. The notice does preserve a right to object.

4. Other notable additions or clarifications

• Support chatbot inputs may be processed in full, not merely the email address.
• Calls may be recorded for marketing purposes.
• Broader descriptions of analytics, cookies, affiliated-company sharing, international transfers, retention, and California privacy rights are included.
• The notice permits retention of aggregated anonymous data indefinitely.

Summary of Available Changes

Scope of the Diff

The diff states only that approximately 5,964 words were removed from the document. It does not identify:

• Which provisions were deleted;
• Whether any provisions were added or replaced;
• The subject matter of the deleted language; or
• Whether the deleted language concerned data use, confidentiality, intellectual property, AI, security, or other topics.

AI Model Training and Customer Data

No specific change regarding the use of customer data to train AI models can be determined from the available diff.

In particular, the diff does not show whether the agreement now:

• Permits or prohibits using customer data to train, fine-tune, or improve AI models;
• Allows use of customer data in aggregated, de-identified, or anonymized form;
• Restricts human review or labeling of customer data;
• Allows retention of customer data for model-development purposes;
• Gives the provider ownership or broad usage rights over inputs, outputs, or derived data;
• Requires consent before customer data is used for AI training; or
• Requires deletion or segregation of customer data from training datasets.

Potential Legal Risks From the Deletion

Although the substance cannot be assessed, removing approximately 5,964 words could create material risk if the deleted provisions included protections such as:

• Confidentiality: Reduced restrictions on disclosure or use of customer information.
• Data-use limitations: Removal of purpose restrictions could expand permitted uses of customer data.
• AI safeguards: Loss of express prohibitions or consent requirements for model training.
• Security obligations: Fewer security controls, incident-notification duties, or audit rights.
• Data ownership: Weaker protections regarding ownership of customer data, inputs, outputs, or derived information.
• Retention and deletion: Less certainty about how long data may be retained or whether it will be deleted.
• Indemnification and liability: Reduced remedies or liability allocation for misuse, breaches, or unauthorized data use.
• Regulatory compliance: Removal of obligations relevant to privacy and data-protection laws.

Conclusion

The supplied diff is insufficient to perform a substantive clause-by-clause analysis. The only confirmed change is a large-scale deletion. The complete redlined text, or at least the deleted and surrounding provisions, is needed to determine whether customer data may now be used for AI training and whether important contractual protections have been removed.