LaunchDarkly

Summary

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

> “Added approximately 52 words to the document”

Accordingly, it is not possible to determine:

• What provisions were added, removed, or replaced;
• Whether the changes alter the parties’ rights or obligations;
• Whether customer data may be used to train, fine-tune, evaluate, or improve AI models;
• Whether data may be shared with affiliates, vendors, or third-party AI providers;
• Whether customer consent, opt-out rights, or deletion rights have changed;
• Whether confidentiality, security, ownership, or liability protections have been modified.

AI-Training Risk Assessment

No conclusion can be reached regarding AI-model training because the actual 52 added words are not included. The added language should be reviewed specifically for terms such as:

• “train,” “fine-tune,” “develop,” “improve,” or “enhance” models;
• “machine learning,” “artificial intelligence,” or “generative AI”;
• “de-identified,” “aggregated,” or “anonymized” data;
• rights to use customer content for “business purposes” or “service improvement”;
• retention or use after termination;
• sharing data with subprocessors or third-party model providers;
• customer opt-out, consent, or deletion mechanisms.

Needed Information

Please provide the actual redline text, using the stated notation—for example, additions in {}, deletions in [], and replacements in []{}. Without that text, no reliable legal-risk analysis can be performed.

Summary of Important Changes

1. Agreement structure and applicable policies

• The introductory text is substantially reorganized into a consolidated “Terms & Policies” presentation.
• The listed policy documents now expressly include, among others:
• Developer Tier and Foundation Self-Service terms
• Invoiced Foundation and Enterprise terms
• Professional Services terms
• Security Program Addendum
• Privacy Policy
• Data Processing Addendum
• Subprocessors
• Support terms and SLAs
• Acceptable Use Policy
• Product-specific terms and archives
• For certain legacy Starter or Pro customers who subscribed before July 7, 2024, the prior terms remain applicable through the end of the applicable Subscription Term.

Risk: Customers must determine which linked policies apply to their plan and whether those policies can be updated independently. The expanded policy list may create additional contractual obligations through incorporated web documents.

2. Updated contracting party and acceptance language

• The agreement identifies the provider as Catamorphic, Co., a Delaware corporation, doing business as LaunchDarkly.
• The acceptance provision is rewritten to state that clicking to accept, accessing, using, or otherwise indicating acceptance of the service binds the customer.
• The agreement applies “to the exclusion of all other terms.”

Risk: Use of the service alone may constitute acceptance, even without a signed order. Customer purchase-order terms or other conflicting terms are expressly subordinated.

3. Data protection and permitted data use

• The definition of Customer Data remains broad: electronic data submitted by or on behalf of the customer, including through third-party services.
• Protection of Customer Data continues to be governed by the Data Processing Addendum (DPA) where legally required.
• The agreement continues to permit LaunchDarkly to:
• Collect, use, and disclose anonymized or aggregated data for business purposes, including industry analysis, benchmarking, and analytics.
• Compile aggregate usage data and use or disclose it.
• The customer remains responsible for the legality, accuracy, quality, and acquisition of Customer Data and for its use within the service.

AI-model training

• No express new provision authorizing LaunchDarkly to use Customer Data to train, fine-tune, or improve artificial-intelligence or machine-learning models appears in the provided diff.
• The existing anonymized/aggregated-data rights could potentially support analytics or benchmarking activities, but the diff does not expressly state that such data may be used for AI training.
• The DPA and Privacy Policy should be reviewed separately for any AI, model-training, de-identification, service-improvement, or subprocessors provisions not shown here.

Risk: “Anonymized” and “aggregated” are not further defined in the displayed language. Customers handling sensitive, confidential, or regulated data should confirm that the DPA limits secondary use and prohibits identifiable Customer Data from being used for model training.

4. Commercial and operational provisions

• Pricing, service plans, fees, support, SLAs, and product-specific terms are more clearly cross-referenced.
• Subscription renewals continue at then-current rates unless cancelled through the account dashboard before the next billing date.
• LaunchDarkly may suspend access for specified risks or breaches, including on notice for certain contractual violations.
• Customer indemnification and liability for breaches of Customer Responsibilities remain significant.

5. Other notable changes

• Notice information is updated to include LaunchDarkly’s Oakland address and separate billing and legal-notice email addresses.
• LaunchDarkly retains authorization to use the customer’s name and logo in marketing materials.
• Numerous edits are grammatical, formatting, or punctuation changes and do not appear legally substantive.

Summary

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

> “Added approximately 52 words to the document”

Because the new wording is not included, it is not possible to determine what legal provisions changed or whether the changes create new risks.

AI Training and Data Use

• No language addressing the customer’s data, artificial intelligence, machine learning, model training, prompts, outputs, or data retention is included in the diff.
• Accordingly, it cannot be determined whether the amendment:
• Permits the provider to use customer data to train or improve AI models;
• Limits training use to aggregated, de-identified, or anonymized data;
• Allows human review or use by subcontractors;
• Grants the customer an opt-out or requires consent;
• Changes ownership or licensing rights in customer inputs or AI-generated outputs;
• Expands data-retention, security, or confidentiality rights; or
• Allows customer data to be transferred to third-party AI providers.

Risk Assessment

The principal issue is that the substantive 52-word addition is missing. Any legal analysis would be speculative without the exact added text and its location in the agreement.

Information Needed

Please provide the actual 52-word addition, including any surrounding text showing where it appears in the contract. This is particularly important if the amendment concerns:

• Data processing or permitted uses;
• Confidentiality;
• Service improvement;
• Artificial intelligence or machine learning;
• Subprocessors or third-party vendors;
• Data retention or deletion; or
• Ownership and licensing of inputs and outputs.

Summary of Important Changes

1. New structure and renamed contracting entity

• The terms are reorganized under a broader “Terms & Policies” framework.
• The agreement now expressly applies to the Developer Tier and Foundation Self-Service offerings.
• The contracting party is changed/clarified from Catamorphic, Co., a Delaware corporation, dba LaunchDarkly to LaunchDarkly in the operative acceptance language. The exact legal-entity relationship should be confirmed, because inconsistent entity naming could create contracting uncertainty.
• The agreement now states that acceptance occurs by clicking to accept, accessing or using the service, or otherwise indicating acceptance.

2. Legacy customers and archived terms

• Customers who entered into a LaunchDarkly Self-Service Starter or Pro subscription before July 7, 2024 are expressly directed to the applicable archived terms at LaunchDarkly’s archive URL until the end of their applicable subscription term.
• This creates a clearer transition rule for legacy customers, but customers should verify which archived terms apply and whether the archive can be changed or removed.
• The agreement now states that it applies “to the exclusion of all other terms.” This may override conflicting purchase-order terms, website terms, or customer procurement documents unless separately accepted by LaunchDarkly.

3. Incorporation of additional policies

The presentation now identifies and links to a wider collection of related policies, including:

• Subscription terms
• Professional services terms
• Security program addendum
• Privacy policy
• Data Processing Addendum
• Subprocessors
• Support terms and SLAs
• Acceptable Use Policy
• Product-specific terms
• Archives

These incorporated documents may contain material obligations, privacy commitments, service limitations, or liability provisions. Customers should review the versions in effect at signing and any update mechanisms.

4. Customer data and AI-model training

• The definition of Customer Data remains broad: it covers electronic data submitted by or on behalf of the customer, including through third-party services.
• The diff does not add an express right to use Customer Data to train, fine-tune, or improve artificial-intelligence or machine-learning models.
• The agreement continues to permit LaunchDarkly to collect, use, and disclose anonymized or aggregated data for business purposes, including industry analysis, benchmarking, and analytics.
• It also permits compilation and use/disclosure of aggregate data concerning customer usage of the service.
• Although these provisions do not expressly authorize AI training, sufficiently anonymized or aggregated information could potentially be used for analytics or model development unless restricted by the DPA, privacy policy, or another incorporated policy. Customers seeking a firm prohibition on AI training should request one expressly.

5. Mostly non-substantive edits

Many changes are punctuation, capitalization, apostrophe, quotation-mark, spacing, URL, or formatting corrections. No major changes are apparent to payment, suspension, warranties, indemnification, liability caps, renewal, termination, or free-trial provisions.

Summary

The provided diff states only that approximately 52 words were added; it does not include the actual added language or identify where it appears in the agreement.

Key Legal Changes

• Cannot be determined: The substance, scope, and legal effect of the additions cannot be analyzed without the new wording.
• No confirmed changes identified: There is insufficient information to determine whether the additions affect:
• Customer data ownership or licensing;
• Permitted uses of customer data;
• Confidentiality or privacy obligations;
• Data retention or deletion;
• Security responsibilities;
• Indemnities, liability, or compliance obligations; or
• AI or machine-learning activities.

AI Training and Model-Use Review

The diff does not provide enough text to determine whether customer data may be:

• Used to train, fine-tune, or improve artificial intelligence or machine-learning models;
• Combined with other customers’ data for model development;
• Reviewed by human personnel for model training or quality assurance;
• Retained after the customer relationship ends for training purposes;
• Used to create outputs, embeddings, derivative datasets, or other model-related materials; or
• Excluded from training but still processed for service improvement, analytics, or safety monitoring.

Risk Assessment

The principal risk is that the actual added terms may contain material changes that are not visible in the supplied diff. In particular, any language granting a broad, perpetual, irrevocable, worldwide, sublicensable, or royalty-free license to customer data could materially expand the provider’s rights, including rights to use data for AI development.

Required Information

To perform a meaningful legal review, provide the actual 52-word addition, preferably together with the surrounding unchanged provisions.

Summary of Important Changes

1. Contracting entity and agreement formation

• The agreement is now presented as being offered by “LaunchDarkly” / Catamorphic, Co., a Delaware corporation, doing business as LaunchDarkly, rather than the prior formulation.
• The acceptance language is reorganized and now expressly states that the customer agrees to the agreement by:
• Clicking to accept;
• Accessing or using the service; or
• Otherwise indicating acceptance.
• The agreement is accepted “to the exclusion of all other terms.”

Risk: Customers may be bound without a separate signature, including through ordinary use of the service. The “exclusion of all other terms” language may conflict with purchase orders, vendor terms, or other customer-provided conditions.

2. Incorporated policies and historical terms

• The introductory policy list has been substantially reorganized and now references multiple documents, including:
• Subscription terms for Foundation and Enterprise;
• Professional Services terms;
• Security Program Addendum;
• Privacy Policy;
• Data Processing Addendum;
• Subprocessor information;
• Support terms and SLAs;
• Acceptable Use Policy;
• Product-specific and other policy documents.
• The prior language specifically addressing customers who entered into certain Starter or Pro subscriptions before July 7, 2024, and directing them to archived terms, appears to have been removed from the main introductory text.

Risk: Customers may now need to determine which combination of online policies applies to them. The removal or relocation of the archived-terms language could create uncertainty about which historical version governs legacy subscriptions.

3. Customer Data and AI-model training

• The definition of “Customer Data” remains broad: it covers electronic data submitted by the customer or on its behalf by a third-party service.
• The diff does not add an express right for LaunchDarkly to use Customer Data to train, fine-tune, or improve artificial-intelligence or machine-learning models.
• The agreement continues to permit LaunchDarkly to:
• Use and disclose anonymized or aggregated data for business purposes, including industry analysis, benchmarking, and analytics; and
• Compile aggregate data about customer usage.
• No clear change was identified that would authorize use of identifiable Customer Data, prompts, feature-flag data, or other customer content for AI training.

Risk: Although there is no express AI-training authorization in this diff, the anonymization and aggregation permissions remain broad. Customers should confirm in the DPA, Privacy Policy, Product-Specific Terms, and any AI-specific terms whether telemetry, usage data, or de-identified data may be used to develop or train models.

4. Notices and administrative provisions

• LaunchDarkly’s notice information is updated to include a physical Oakland address and separate billing and legal-notice email addresses.
• Most other changes are grammatical, punctuation, capitalization, or typographical corrections and do not appear to alter substantive rights.

Overall assessment

The principal legal changes concern contract formation, identification of the contracting entity, incorporation of online policies, and possible uncertainty for legacy customers. No new express customer-data AI-training right appears in the provided diff, but the existing anonymized/aggregated-data rights warrant review alongside the referenced privacy and data-processing documents.

Summary of Important Changes

1. Contracting entity and document structure

• The agreement is now presented as LaunchDarkly “Terms & Policies” / Terms of Service rather than primarily as terms offered by Catamorphic, Co., d/b/a LaunchDarkly.
• The contracting entity appears to be identified as LaunchDarkly, a Delaware corporation, with the document referring to Developer Tier and Foundation Self-Service offerings.
• The introductory policy list has been expanded or reorganized to reference several separate documents, including:
• Subscription Terms
• Professional Services Terms
• Security Program Addendum
• Privacy Policy
• Data Processing Addendum
• Subprocessors
• Code of Conduct
• Support Terms and SLAs
• Acceptable Use Policy
• Product-specific or partner terms
• These linked documents may become part of the contractual framework. Customers should verify which documents apply to their service and review their version-control and amendment provisions.

2. Legacy customer treatment

• A specific transition rule is added for customers who entered into a LaunchDarkly Self-Service Starter or Pro subscription before July 7, 2024.
• Their use of the services remains governed by the applicable archived terms at LaunchDarkly’s archives URL until the end of the applicable Subscription Term.
• This creates different contractual regimes for legacy and newer customers. Customers should preserve copies of the archived terms and confirm which version governs renewals, amendments, and post-renewal use.

3. Acceptance and priority of terms

• The revised language expressly states that clicking to accept, accessing, using, or otherwise indicating acceptance binds the customer to the agreement.
• The agreement applies “to the exclusion of all other terms.” This may limit the effectiveness of purchase-order terms, procurement portals, or customer-standard terms unless expressly accepted by LaunchDarkly.
• The customer must represent that the accepting person has authority to bind the relevant organization.

4. Customer data and privacy

• The definitions and data-protection provisions are largely retained, including:
• Customer Data: electronic data submitted by or for the customer, including through third-party services.
• Customer responsibility for the legality, accuracy, quality, and collection of Customer Data.
• Processing of personal data under the Data Processing Addendum.
• Restrictions concerning sensitive personal data.
• The agreement continues to permit LaunchDarkly to collect, use, and disclose anonymized or aggregated data for business purposes, including industry analysis, benchmarking, and analytics.
• A separate confidentiality provision also permits compilation and use or disclosure of aggregate usage data.

5. AI model training

• No express new authorization to use Customer Data to train artificial-intelligence or machine-learning models appears in the supplied diff.
• The existing permissions for anonymized, aggregated, and usage data could potentially cover analytics or model development depending on how “anonymized,” “aggregated,” and “business purposes” are interpreted in the complete agreement and Privacy Policy.
• Customers should confirm in the Privacy Policy, DPA, product-specific terms, and any AI terms whether:
• Customer Data or prompts may be used for model training;
• de-identified data may be used for training;
• customers can opt out; and
• customer content is excluded from generalized or third-party models.

6. Other notable changes

• Administrative wording, punctuation, capitalization, and contact details have been updated.
• The customer’s name and logo authorization remains.
• Most operational, payment, warranty, liability, suspension, termination, and free-trial provisions appear substantively unchanged in the supplied diff.

Summary

The provided diff does not include the actual 52 added words. It only states that approximately 52 words were added.

Legal and AI-Training Analysis

• No substantive changes can be identified without seeing the added language.
• It is therefore not possible to determine whether the amendment:
• Permits or restricts using customer data to train, fine-tune, or improve AI models;
• Allows use of customer data for generalized model training or only for providing the contracted services;
• Treats prompts, inputs, outputs, usage data, or metadata as customer data;
• Allows the provider to retain customer data after termination;
• Permits de-identification, aggregation, or anonymization of customer data;
• Gives the provider rights to use customer content for product development, analytics, or marketing;
• Changes confidentiality, security, deletion, or data-processing obligations; or
• Transfers responsibility or liability for AI-generated outputs to the customer.

Risk Assessment

The absence of the actual text prevents a meaningful assessment of new legal risks. In particular, any statement that customer data may be used to “improve services,” “develop models,” or for “research” could create broad permission for AI training unless limited by clear restrictions.

Please provide the actual inserted language, including any surrounding clause or the complete redline, for a reliable legal analysis.

Summary of Important Changes

1. New or reorganized terms framework

• The opening language changes from a reference to archived terms applying through the subscription term to a more detailed “Terms & Policies” structure.
• The revised policy list expressly identifies multiple documents, including:
• Subscription Terms
• Professional Services Terms
• Security Program Addendum
• Privacy Policy
• Data Processing Addendum
• Subprocessor information
• Support terms and SLAs
• Acceptable Use Policy
• Product-specific terms
• Risk: More external documents are incorporated into the contractual framework. Customers should confirm which documents apply to their service, plan, order, and customer type, and whether those documents can be changed unilaterally.

2. Updated contracting party and agreement language

• The agreement is revised to identify the provider as Catamorphic, Co., a Delaware corporation, doing business as LaunchDarkly.
• The acceptance language is reorganized and now expressly states that the agreement applies when the customer:
• Clicks to accept;
• Accesses or uses the service; or
• Otherwise indicates acceptance.
• The agreement is accepted “to the exclusion of all other terms.”
• Risk: This may make purchase-order terms, procurement terms, or other customer terms ineffective unless expressly accepted by LaunchDarkly.

3. Customer Data and data protection

• “Customer Data” remains defined broadly as electronic data submitted by or for the customer, including through third-party services.
• The agreement continues to state that customer data protection is governed, where applicable, by the Data Processing Addendum (DPA).
• Customer remains responsible for the accuracy, legality, collection methods, and use of Customer Data.
• LaunchDarkly may use anonymized or aggregated data for business purposes, including industry analysis, benchmarking, and analytics. A similar aggregate-usage provision also appears in the confidentiality section.

4. AI model training

• No express change concerning use of Customer Data to train AI models appears in this diff.
• The revised text does not add an explicit authorization to use Customer Data, prompts, outputs, telemetry, or other customer content to train, fine-tune, evaluate, or improve artificial-intelligence models.
• Conversely, the diff does not add an express prohibition on AI training. Any such rights or restrictions may exist in the Privacy Policy, DPA, Product-Specific Terms, or another incorporated policy.
• Recommended review: Check all incorporated policies for definitions of “Service Data,” “Usage Data,” “Aggregated Data,” “Customer Data,” and any AI-training or model-improvement rights.

5. Other notable changes

• Numerous edits are grammatical, punctuation, capitalization, or formatting changes and do not appear substantively important.
• Notice information is reformatted and adds LaunchDarkly billing and legal-notice email addresses.
• LaunchDarkly retains authorization to use the customer’s name and logo in marketing materials.

Summary

The diff indicates that approximately 52 words were added, but the actual added wording is not included. Because no clause text is available, the legal and commercial impact cannot be assessed reliably.

AI Training and Data Use

• No determination can be made about whether the new language:
• Authorizes the provider to use customer data to train, fine-tune, or improve AI models;
• Limits such use to aggregated, anonymized, or de-identified data;
• Permits use of customer prompts, inputs, outputs, files, or personal information;
• Applies to human review or other processing for model development;
• Gives the customer an opt-out or requires separate consent; or
• Allows data to be retained after termination.

Potential Risks Requiring Review

If the added language concerns AI or data use, key risks may include:

• Broad license rights: The provider may receive expansive rights to copy, analyze, modify, or create derivative works from customer data.
• Confidentiality exposure: Customer confidential information or personal data could be used in model development or exposed through model outputs.
• Lack of limitations: The clause may not restrict the purposes, duration, recipients, or geographic scope of data use.
• No deletion commitment: Data may remain in training datasets, backups, logs, or derived models after termination.
• Regulatory and contractual conflicts: AI training may conflict with privacy laws, data-processing agreements, confidentiality obligations, or sector-specific requirements.
• Insufficient safeguards: The language may omit security requirements, de-identification standards, audit rights, or breach obligations.

Conclusion

The diff is incomplete and contains only a description of the number of added words, not the words themselves. The substantive changes—and particularly any changes regarding AI model training or customer-data usage—cannot be identified without the actual added text.

Summary

The diff only states that approximately 52 words were added, but does not provide the actual added language or identify where it appears.

Key Legal Implications

• No substantive comparison is possible. Without the text of the addition, it is not possible to determine whether the changes affect:
• Customer data ownership or licensing
• Privacy or confidentiality obligations
• Data retention or deletion
• Disclosure to third parties
• Service providers or subprocessors
• Artificial intelligence or machine-learning activities
• Liability, indemnification, or regulatory compliance

• AI-training provisions cannot be assessed. The provided diff does not indicate whether customer data may be:
• Used to train, fine-tune, or improve AI or machine-learning models
• Used in aggregated, de-identified, or anonymized form
• Retained after termination for model-training purposes
• Shared with AI vendors or other third parties
• Excluded from training unless the customer opts in or opts out

• Potential review risk. A short addition can still materially expand the provider’s rights, particularly if it introduces broad language such as “improve the services,” “develop models,” “analyze data,” or “use aggregated information.” The actual wording is necessary to evaluate whether such rights include customer content, personal data, confidential information, or derived data.

Conclusion

No reliable legal or AI-data-use conclusions can be drawn from the supplied diff. The approximately 52 new words should be provided in full, together with their location in the agreement, for a meaningful risk analysis.

Structured Summary of Important Changes

1. Contracting entity and branding

• The agreement is revised from being offered by Catamorphic, Co., a Delaware corporation, doing business as LaunchDarkly to being offered by LaunchDarkly, Inc. or a similarly branded LaunchDarkly entity.
• The document heading and policy-navigation language are substantially reorganized under “Terms & Policies.”
• Risk: Customers should confirm the correct legal contracting entity, especially for governing law, enforcement, invoicing, data-processing obligations, and parent/subsidiary liability.

2. Acceptance and contract formation

• The introductory language is reorganized and expressly states that customers accept the agreement by:
• clicking to accept;
• accessing or using the service; or
• otherwise indicating acceptance.
• The revised text states that the agreement applies “to the exclusion of all other terms.”
• Risk: This may make customer purchase-order terms, procurement terms, or other conflicting documents less likely to apply. Customers should verify whether an Order, master agreement, or negotiated terms expressly take precedence.

3. Incorporated policies and archived terms

• The prior language referred certain customers—particularly older Starter or Pro subscribers—to applicable terms in LaunchDarkly’s archives until the end of the applicable subscription term.
• The revised presentation appears to list multiple policies and addenda, including subscription terms, professional-services terms, security materials, the Privacy Policy, DPA, subprocessors, support terms, SLA, Acceptable Use Policy, and product-specific terms.
• Risk: More external documents may be incorporated into the contractual framework. Their contents and amendment mechanisms should be reviewed, including whether LaunchDarkly can update them unilaterally.

4. Customer Data and privacy

• The definition of “Customer Data” remains data submitted by or for the customer to the service.
• The agreement continues to state that protection of personal data is governed by the Data Processing Addendum (DPA).
• Customer remains responsible for the legality, accuracy, quality, acquisition, and use of Customer Data.
• LaunchDarkly may collect, use, disclose, and compile anonymized or aggregated data for business purposes, including industry analysis, benchmarking, analytics, and usage analysis.

5. AI-model training

• No express new authorization to use Customer Data to train AI models appears in the provided diff.
• The existing anonymized/aggregated-data rights could potentially cover analytics or benchmarking involving service-derived information, but the excerpt does not expressly mention:
• artificial intelligence;
• machine learning;
• model training or fine-tuning;
• generative AI; or
• use of identifiable Customer Data for training.
• Customers should seek confirmation that Customer Data, prompts, outputs, configurations, and telemetry will not be used to train or improve AI models unless expressly authorized.

6. Other changes

• Most remaining changes appear to be punctuation, capitalization, apostrophe, formatting, or document-navigation corrections.
• No material change to liability, indemnity, suspension, payment, termination, or warranty provisions is clearly shown in this diff.

Summary of Important Changes

1. Contract structure and incorporated policies

• The introductory language has been substantially reorganized and now presents a consolidated “Terms & Policies” framework, including:
• Terms of Service
• Developer Tier and Foundation Self-Service terms
• Invoiced Foundation and Enterprise terms
• Professional Services terms
• Security Program Addendum
• Privacy Policy
• Data Processing Addendum
• Subprocessor information
• Support terms and SLAs
• Acceptable Use Policy
• Product-specific and other policy documents

• The prior language specifically addressing customers who subscribed to LaunchDarkly Starter or Pro plans before July 7, 2024 is moved into the broader policy structure. It continues to state that those customers’ use is governed by the applicable archived terms at the policy archive URL through the end of the applicable Subscription Term.

• The agreement now more clearly states that customers accept the Terms by clicking to accept, accessing or using the service, or otherwise indicating acceptance. It also expressly states that the agreement applies “to the exclusion of all other terms.”

Risk

• Customers may be bound by multiple online policies incorporated by reference. Those documents may be updated separately, so customers should review the incorporated policies and any change provisions carefully.
• The exclusion of other terms may create a risk that conflicting purchase-order terms or customer-provided terms will be rejected.

2. Contracting entity and terminology

• The opening provisions identify the provider as Catamorphic, Co., a Delaware corporation, doing business as LaunchDarkly, rather than using only “LaunchDarkly” or “Company.”
• Definitions have been cleaned up and standardized, including Customer Data, Customer Equipment, Documentation, Entitlements, Services, Fees, Orders, Sensitive Data, and Third-Party Services.
• The definition of Customer Data continues to cover electronic data submitted by or on behalf of the customer, including through third-party services.

3. Customer Data and AI-model training

• No express new authorization to use Customer Data to train artificial intelligence or machine-learning models appears in this diff.
• The provisions shown continue to address:
• Protection of Customer Data under the agreement and the Data Processing Addendum.
• Customer responsibility for the legality, accuracy, quality, and acquisition of Customer Data.
• LaunchDarkly’s ability to use or disclose anonymized or aggregated data for business purposes, including industry analysis, benchmarking, and analytics.
• Use of aggregate usage data in connection with providing and improving the service.

AI-related risk assessment

• Although no AI-training language is added here, the existing anonymized/aggregated-data rights could potentially permit use of derived data in analytics or product-development activities, depending on how “anonymized” and “aggregated” are defined elsewhere.
• Customers seeking to prohibit model training should confirm whether the Privacy Policy, DPA, Product-Specific Terms, or other incorporated policies contain separate AI-training provisions.

Summary

The provided diff states only that “approximately 52 words” were added. It does not include the actual added language or identify any deletions or replacements.

Key Legal Changes

• Cannot be determined: The substantive changes, affected obligations, and allocation of risk cannot be analyzed without the text of the 52 added words.
• No identifiable revisions: The diff does not show any specific changes to warranties, liability, confidentiality, data rights, termination, security, or intellectual-property provisions.

AI Training and Customer Data

• No determination possible: The provided information does not reveal whether the customer’s data may be:
• Used to train, fine-tune, or improve artificial-intelligence or machine-learning models;
• Combined with other customers’ data;
• Reviewed by human personnel for model development or quality assurance;
• Retained after termination for training or analytics; or
• Used in de-identified, aggregated, or identifiable form.
• Any new language granting the provider rights to use customer content, prompts, inputs, outputs, usage data, or metadata for “improvement,” “research,” “analytics,” or “model development” could materially expand the customer’s confidentiality, privacy, intellectual-property, and regulatory risks.

Recommended Next Step

Please provide the actual redlined language, including the 52 added words and any surrounding provisions. The AI-training analysis should specifically assess:

1. What categories of customer data may be used;

2. Whether use for AI training is opt-in, opt-out, or automatic;

3. Whether data is de-identified and whether re-identification is prohibited;

4. Whether the provider may share data with affiliates or subprocessors;

5. How long data is retained;

6. Whether the customer owns resulting models or improvements; and

7. Whether the provider provides a contractual commitment not to train on customer data.

Structured Summary of Important Changes

1. Contracting entity and agreement formation

• The terms now identify the provider as “Catamorphic, Co., a Delaware corporation, dba as LaunchDarkly” and refer to it as “LaunchDarkly” or “Company.”
• The agreement is expressly accepted by:
• Clicking to accept;
• Accessing or using the service in any manner; or
• Otherwise indicating acceptance.
• The customer agrees to be bound by the agreement “to the exclusion of all other terms.”

Risk: This strengthens a “clickwrap/browsewrap” formation theory and may make purchase orders, customer terms, or other conflicting terms ineffective. Customers should confirm that authorized personnel accept the terms and review purchase-order precedence provisions, if any.

2. Broader and reorganized policy framework

• The opening text now refers to a consolidated “Terms & Policies” structure, including:
• Developer Tier and Foundation Self-Service terms;
• Foundation invoiced and Enterprise terms;
• Professional services terms;
• Security, privacy, data-processing, support, SLA, acceptable-use, subprocessors, and other policies.
• The revised text appears to incorporate multiple linked policies and product-specific terms by reference.
• The prior language specifically addressing Starter or Pro subscriptions entered into before July 7, 2024 is retained in substantially similar form, including application of archived terms until the end of the applicable subscription term.

Risk: Customers may be bound by policies located on separate webpages, including policies that may be updated independently. The applicable version and hierarchy among these documents should be confirmed.

3. Data-use and AI-model training

• The definition of Customer Data remains substantially unchanged: electronic data or information submitted by or for the customer.
• The provisions reviewed continue to state that LaunchDarkly may:
• Use and disclose anonymized or aggregated data for business purposes, including industry analysis, benchmarking, and analytics; and
• Compile aggregate data concerning customer usage and use or disclose that aggregate data.
• No express new authorization to use Customer Data to train AI or machine-learning models appears in this diff.
• Conversely, the diff does not add an express prohibition on using Customer Data, prompts, outputs, usage data, or derived data for AI training.

Risk: Although no direct AI-training permission is added, the broad anonymized/aggregated-data language could potentially cover data-derived analytics or model-improvement activities if the data is sufficiently deidentified. Customers requiring a strict no-training commitment should seek explicit language covering Customer Data, confidential information, prompts, outputs, telemetry, and derived data.

4. Other notable changes

• Several changes are editorial or punctuation/formatting corrections, including apostrophes, quotation marks, and defined-term formatting.
• Notice information is reorganized, adding LaunchDarkly’s Oakland address and separate billing and legal-notice email addresses.
• The customer’s name and logo authorization remains present.
• No material changes are apparent in the suspension, payment, indemnity, liability, termination, or free-trial provisions based on the supplied diff.

Summary

Scope of the Change

• The diff states that approximately 52 words were added to the document.
• However, it does not provide the actual added language or identify where the additions appear.

Legal and Commercial Impact

• The legal effect of the changes cannot be reliably assessed without reviewing the specific wording.
• The additions could affect important areas such as:
• Customer data ownership and permitted use
• Confidentiality and privacy obligations
• Service-provider rights to analyze or share data
• Liability, indemnification, or regulatory compliance
• Termination and post-termination data deletion
• Intellectual-property rights in outputs or derived data

AI Training and Model Development

• The supplied diff contains no substantive language addressing AI, machine learning, model training, model improvement, or use of customer data.
• Accordingly, it is not possible to determine whether the revised document:
• Permits customer data to train or fine-tune AI models
• Allows use of data to improve products or services
• Permits retention of prompts, inputs, outputs, or derived data
• Allows data to be shared with affiliates, subprocessors, or third-party model providers
• Provides opt-out, deletion, confidentiality, or anonymization protections
• Grants the provider rights to use aggregated or de-identified data

Key Risk Assessment

• The principal risk is insufficient information, rather than an identifiable contractual change.
• A word-count statement alone cannot establish whether customer data may be used for AI training or whether existing protections were narrowed.

Information Needed

To complete the analysis, provide the actual 52-word addition and, preferably, the surrounding original and revised text. Particular attention should be given to terms such as “customer data,” “content,” “inputs,” “outputs,” “improve,” “develop,” “train,” “fine-tune,” “machine learning,” “artificial intelligence,” “aggregated,” “de-identified,” and “service providers.”

Diff Analysis

Available Information

The diff only states:

> “Added approximately 52 words to the document”

It does not include the actual added language, deleted language, or replacement text.

Important Changes and Risks

Because the substantive wording is missing, it is not possible to determine:

• What contractual obligations have changed;
• Whether liability, indemnification, warranties, termination, confidentiality, or dispute-resolution provisions were revised;
• Whether the customer’s rights or the provider’s rights have expanded;
• Whether any new fees, renewal terms, restrictions, or data-security obligations were added; or
• Whether the changes create new legal or commercial risks.

AI Training and Customer Data

The available diff does not identify whether customer data may be used to train, fine-tune, evaluate, or improve artificial-intelligence models.

The missing language should be reviewed specifically for terms such as:

• “train,” “fine-tune,” “improve,” or “develop” models or services;
• “machine learning,” “artificial intelligence,” or “generative AI”;
• rights to use customer content, prompts, inputs, outputs, usage data, or telemetry;
• broad licenses allowing use of data for product development or analytics;
• de-identification, aggregation, or pseudonymization provisions;
• opt-out, deletion, retention, or human-review rights; and
• restrictions on using confidential, personal, regulated, or proprietary information.

If the added language grants model-training rights, important risks may include loss of confidentiality, reuse of proprietary information, exposure of personal data, inability to require deletion from trained models, and use of customer data to benefit other customers or the provider.

Conclusion

No reliable legal analysis can be performed from the word-count statement alone. The full redlined text—or at least the 52 added words and any surrounding provisions—is required to identify the actual changes and assess AI-training or data-use implications.

Summary of Important Changes

1. Contracting party and acceptance mechanics

• The terms now identify the provider as Catamorphic, Co., a Delaware corporation, doing business as LaunchDarkly, rather than using the prior abbreviated or differently formatted company description.
• The agreement expressly states that customers accept the terms by:
• Clicking to accept;
• Accessing or using the LaunchDarkly service in any manner; or
• Otherwise indicating acceptance.
• The agreement is stated to apply “to the exclusion of all other terms.” This may reject conflicting purchase orders, procurement terms, or other customer terms unless separately agreed in writing.
• The language now more clearly states that the terms affect the customer’s legal rights and that the accepting person represents authority to bind the relevant organization.

2. Applicable policies and incorporated documents

• The introductory policy list has been reorganized and expanded or clarified to include documents such as:
• Foundation Self-Service terms;
• Invoiced and Enterprise terms;
• Professional Services terms;
• Security Program Addendum;
• Privacy Policy;
• Data Processing Addendum;
• Subprocessors information;
• Support terms and SLAs;
• Acceptable Use Policy;
• Product-specific and other policy documents.
• The prior language referring customers to archived terms until the end of the applicable subscription term appears to have been replaced with a consolidated current policy structure. Customers should confirm which version of each linked document applies and whether LaunchDarkly may update those documents unilaterally.
• Several URLs and incorporated policies remain legally significant because they may impose additional obligations beyond the main agreement.

3. Customer Data and privacy

• The definition of Customer Data remains broad: electronic data submitted by the customer or on its behalf through a third-party service.
• The agreement continues to place responsibility on the customer for the legality, accuracy, quality, acquisition, and use of Customer Data.
• Protection of Customer Data remains tied to the Data Processing Addendum (DPA), which should be reviewed for processing instructions, subprocessors, international transfers, security obligations, and deletion/return requirements.
• The agreement continues to permit LaunchDarkly to collect, use, and disclose anonymized or aggregated data for business purposes, including industry analysis, benchmarking, and analytics. The confidentiality provisions also permit aggregate usage data to be compiled and used or disclosed.

4. AI model training

• No express new provision authorizing or prohibiting the use of Customer Data to train AI models appears in this diff.
• The retained permissions for anonymized or aggregated data could potentially cover analytics or model-development activities if data is genuinely anonymized or aggregated, but the diff does not expressly say that Customer Data may be used to train AI systems.
• Customers should seek clarification on whether prompts, inputs, outputs, telemetry, or de-identified data may be used for AI training, and whether the DPA or product-specific terms contain additional language.

5. Other notable points

• Most remaining changes are grammatical, punctuation, capitalization, or formatting corrections.
• Existing customer obligations, indemnities, suspension rights, liability limitations, automatic renewal, and free-trial disclaimers appear substantially unchanged.

Summary

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

> “Added approximately 52 words to the document”

Accordingly, it is not possible to determine:

• What contractual rights or obligations changed;
• Whether liability, confidentiality, security, intellectual property, termination, or payment terms were amended;
• Whether the customer’s data may be used to train, fine-tune, evaluate, or improve AI models;
• Whether such use is limited to anonymized, aggregated, de-identified, or non-personal data;
• Whether the customer can opt out of AI training or withdraw consent;
• Whether the provider may share data with affiliates, subprocessors, or third-party AI vendors;
• Whether the provider claims ownership or broad usage rights in customer data or outputs; or
• Whether any additional privacy, regulatory, or data-security risks were introduced.

AI-Training Risk Assessment

No conclusion can be reached regarding AI-model training because the added 52 words are not provided. The key language to review would include terms such as “train,” “fine-tune,” “improve,” “develop,” “machine learning,” “artificial intelligence,” “service improvement,” “de-identified,” “aggregated,” or “usage data.”

Required Information

Please provide the actual 52-word addition, including any surrounding sentence or paragraph. The precise wording is necessary to assess whether the change:

1. Permits training on customer content by default;

2. Applies to personal data, confidential information, prompts, outputs, or metadata;

3. Allows retention after termination;

4. Permits disclosure to third parties; or

5. Provides meaningful restrictions, deletion rights, or an opt-out mechanism.

Summary of Important Changes

1. Contract structure and incorporated policies

• The introductory text was substantially reorganized and now presents a broader “Terms & Policies” framework, including:
• Developer Tier and Foundation Self-Service terms
• Invoiced Foundation and Enterprise terms
• Professional Services terms
• Security Program Addendum
• Privacy Policy
• Data Processing Addendum
• Subprocessors, Acceptable Use Policy, Support Terms, SLAs, and other policies
• The prior language directed certain customers to archived terms at launchdarkly.com/policies/archives/ through the end of the applicable Subscription Term. That concept appears to remain, but the surrounding policy index has been expanded or reformatted.
• Risk: Customers may be subject to multiple online documents, some of which may be incorporated by reference. The applicable version and order of precedence should be confirmed, particularly where the Agreement conflicts with a DPA, product-specific terms, or archived terms.

2. Contracting entity and acceptance mechanics

• The agreement now identifies the provider as Catamorphic, Co., a Delaware corporation, doing business as LaunchDarkly.
• The customer is expressly bound by accepting the terms, accessing or using the service, or otherwise indicating acceptance.
• New or clarified language states that the Agreement applies “to the exclusion of all other terms.”
• Risk: Use of the service alone may constitute acceptance, even without a signed agreement. The exclusion-of-other-terms provision may reject purchase-order terms, customer policies, or negotiated conditions unless expressly accepted by LaunchDarkly.

3. Customer Data and data protection

• The definition of “Customer Data” remains broad, covering electronic information submitted by or on behalf of the customer, including through third-party services.
• The DPA continues to govern personal data processing where applicable.
• The customer remains responsible for the legality, accuracy, quality, and means of obtaining Customer Data, as well as its use with the services.
• Risk: Customers should ensure they have all required notices, consents, and legal bases for data submitted to the service, particularly sensitive personal data.

4. AI model training and data use

• No express new provision authorizing LaunchDarkly to use Customer Data to train AI models was identified in this diff.
• The agreement continues to permit use or disclosure of:
• Anonymized or aggregated data for business purposes, including industry analysis, benchmarking, and analytics.
• Aggregate data concerning customer usage of the service.
• These permissions appear limited to anonymized, aggregated, or usage-related information rather than identifiable Customer Data. However, the text does not expressly state that such data will not be used for AI training.
• Risk: If AI-training restrictions are important, customers should seek an explicit contractual statement prohibiting use of Customer Data, prompts, configurations, logs, or outputs to train general-purpose or third-party AI models, and clarifying whether anonymized or aggregated data may be used for that purpose.

5. Other changes

• Most remaining edits are grammatical, punctuation, capitalization, or formatting changes and do not appear to materially alter commercial rights.
• Notice information was updated to provide LaunchDarkly’s physical address and separate billing and legal-notice email addresses.

Summary

The supplied diff does not include the actual added or deleted contract language. It only states that approximately 52 words were added.

AI Training and Data-Use Changes

• No determination is possible regarding whether the customer’s data may now be used to train, fine-tune, evaluate, or improve AI models.
• The diff does not reveal whether it changes:
• Ownership or control of customer data;
• The provider’s rights to use customer content;
• Use of data for model training or product improvement;
• Whether data is aggregated, anonymized, or de-identified first;
• Opt-out or consent requirements;
• Retention, deletion, or audit rights;
• Sharing with affiliates, contractors, or third-party AI providers; or
• Confidentiality and security protections.

Legal-Risk Assessment

Because the substantive wording is missing, no reliable assessment can be made of new obligations, expanded provider rights, limitations of liability, confidentiality changes, or privacy risks.

Information Needed

Please provide the actual redline text, using the stated format—for example:

• Additions: {new language}
• Deletions: [deleted language]
• Replacements: [old language]{new language}

The complete diff is necessary to identify material changes and any specific rights to use customer data for AI model training.

Executive Summary

The diff appears to reorganize and republish LaunchDarkly’s terms rather than make broad substantive changes. However, several legal and operational points are important.

Important Changes

1. New consolidated terms structure

The introduction now identifies a broader set of documents and policies, including:

• Developer Tier and Foundation Self-Service terms
• Foundation invoiced and Enterprise terms
• Professional Services terms
• Security Program Addendum
• Privacy Policy and Data Processing Addendum
• Support terms and SLAs
• Acceptable Use Policy
• Product-specific terms and archives

Risk: Customers may be bound by multiple web-based documents incorporated by reference. The terms should be reviewed together, particularly for conflicts among the main agreement, the DPA, privacy policy, support terms, and product-specific terms.

2. Legacy Starter and Pro customers

The revised language expressly states that customers who entered into a LaunchDarkly Self-Service Starter or Pro subscription before July 7, 2024 remain governed by the applicable archived terms until the end of their applicable subscription term.

Effect: This preserves older contractual terms for legacy customers temporarily, but customers may transition to current terms at renewal or under a new order.

3. Expanded acceptance and exclusivity language

The agreement now more clearly states that acceptance may occur by:

• Clicking to accept
• Accessing or using the service
• Otherwise indicating acceptance

It also states that the agreement applies “to the exclusion of all other terms.”

Risk: This may exclude customer purchase-order terms, procurement conditions, or other contradictory terms unless expressly accepted by LaunchDarkly.

4. Contracting entity clarification

The terms identify Catamorphic, Co., a Delaware corporation, doing business as LaunchDarkly, as the contracting company.

Effect: This clarifies the legal entity, which may affect contracting, litigation, invoicing, and enforcement analysis.

5. Notice information updated

The notice provision now lists LaunchDarkly’s physical address and separate billing and legal-notice email addresses.

Effect: Legal notices may need to be sent to the newly specified addresses to be effective.

Customer Data and AI Training

No new provision expressly authorizing LaunchDarkly to use Customer Data to train artificial-intelligence or machine-learning models appears in the supplied diff.

The language continues to provide that:

• Customer Data is information submitted by or for the customer.
• LaunchDarkly must protect Customer Data under the agreement and DPA.
• LaunchDarkly may collect, use, and disclose anonymized or aggregated data for business purposes, including industry analysis, benchmarking, and analytics.
• LaunchDarkly may compile aggregate data regarding customer use of the service.

Risk: Although this is not an express AI-training grant, broad “business purposes,” analytics, anonymization, and aggregation rights could potentially support model development if data is transformed into genuinely anonymized or aggregated information. The DPA and Privacy Policy should be checked for any separate AI-training, service-improvement, de-identification, or model-development language.

Summary of Important Changes

1. Expanded and reorganized policy framework

• The introductory text replaces a narrow reference to “Terms of Service” with a broader “Terms & Policies” framework.
• The policy list now expressly identifies multiple documents, including:
• Subscription Terms
• Professional Services Terms
• Security Program Addendum
• Privacy Policy
• Data Processing Addendum
• Subprocessor information
• Acceptable Use Policy
• Support Terms and SLAs
• Product-specific and other program terms
• The agreement appears to incorporate or point users to additional policies through linked webpages. This may make the customer subject to obligations outside the main agreement, potentially including documents that LaunchDarkly can update separately.

Risk: Customers should confirm which linked policies are contractually incorporated, how conflicts are resolved, and whether LaunchDarkly may amend them unilaterally.

2. Revised contracting entity and agreement acceptance language

• The contracting party is clarified as Catamorphic, Co., a Delaware corporation, doing business as LaunchDarkly.
• The acceptance language is revised to state that clicking to accept, accessing, using, or otherwise indicating acceptance binds the customer.
• The agreement is accepted “to the exclusion of all other terms.”

Risk: Customer purchase orders, procurement terms, or other standard terms may be expressly rejected, even if referenced in an order or invoice. The expanded “acceptance by use” language may also bind users without a separate signature.

3. Customer Data and privacy terms

• “Customer Data” remains broadly defined as electronic data submitted by the customer or on its behalf through a third-party service.
• Data protection obligations continue to be addressed through the Data Processing Addendum, rather than materially expanded in the main agreement.
• The customer remains responsible for the accuracy, legality, collection, and use of Customer Data.

4. AI-model training and data use

• No express new authorization to use Customer Data to train AI models appears in this diff.
• The existing provisions permit LaunchDarkly to:
• Collect, use, and disclose anonymized or aggregated data for business purposes, including industry analysis, benchmarking, and analytics.
• Compile aggregate data concerning customer use of the service and use or disclose that data.
• These provisions do not expressly state whether data used for analytics may also be used to train, fine-tune, evaluate, or improve AI or machine-learning models.

Risk: Although no direct AI-training permission is added, “business purposes,” “analytics,” and “aggregate data” could potentially be interpreted broadly. Customers seeking certainty should request an explicit contractual statement that Customer Data, prompts, inputs, outputs, and usage data will not be used to train generative AI or other models, except as specifically authorized.

5. Other notable changes

• Numerous changes are editorial or typographical, including punctuation, capitalization, apostrophes, and formatting.
• Notice information is reorganized and provides separate LaunchDarkly billing and legal-notice addresses.
• Existing rights regarding suspension, payment, indemnification, confidentiality, liability limitations, renewal, and termination appear substantively unchanged in the supplied diff.

Summary

The diff does not include the text of the approximately 52 added words. It only states that words were added.

Legal and Risk Analysis

• Substantive changes: Cannot be determined without the actual added language.
• Customer data use: No conclusion can be reached about whether the additions:
• Permit use of customer data to train, fine-tune, or improve AI models;
• Allow data to be shared with affiliates, vendors, or third-party AI providers;
• Expand the definition of “customer data” or authorize de-identification or aggregation;
• Change consent, opt-out, deletion, confidentiality, or retention rights; or
• Permit use of prompts, inputs, outputs, usage data, or personal information for model development.
• New legal risks: Cannot be assessed from the description alone. The added language could materially affect data ownership, confidentiality, privacy compliance, intellectual-property rights, or liability, but the specific impact depends on its wording.

Information Needed

Please provide the actual 52-word addition, using the stated markup format. Without the text of the addition and its surrounding provisions, a reliable contract analysis is not possible.

Summary

The provided diff only states:

> “Added approximately 52 words to the document”

It does not include the actual added, deleted, or replaced contract language. As a result, the legal effect of the changes cannot be determined.

AI Training and Customer Data

• No substantive language concerning customer data, use of data, AI model training, machine learning, service improvement, or data retention is included.
• It is therefore not possible to determine whether the changes:
• Permit or prohibit training AI models on customer data;
• Allow use of customer content for model improvement or product development;
• Distinguish between personal data, prompts, outputs, telemetry, or de-identified data;
• Require customer consent or provide an opt-out;
• Assign ownership of training inputs or resulting models;
• Impose confidentiality, deletion, security, or data-localization obligations.

Risk Assessment

Because the 52 added words are not provided, no reliable assessment can be made of:

• New permissions granted to the provider;
• Expanded data-use rights;
• Changes to confidentiality or privacy obligations;
• Changes to liability, indemnity, or compliance responsibilities;
• Whether existing restrictions were narrowed or removed.

Required Information

To perform the requested analysis, provide the actual redlined text showing the 52-word addition, including any surrounding provisions necessary to understand its meaning.

Structured Summary of Important Changes

1. New and reorganized policy framework

The opening section has been substantially rewritten and now presents a consolidated “Terms & Policies” framework, including links or references to:

• Developer Tier and Foundation Self-Service terms
• Foundation (Invoiced) and Enterprise subscription terms
• Professional Services terms
• Application and Platform Readiness terms
• Security Program Addendum
• Privacy Policy and Data Processing Addendum
• Subprocessors, Code of Conduct, Support Terms, SLAs, Partner Program, Integrations, Acceptable Use Policy, VPAT, and Archives

Risk: Customers may now be subject to a broader set of incorporated policies. The applicable policy should be identified for the particular product, plan, and order.

2. Legacy terms may apply only temporarily

The revised language states that certain older Starter or Pro customers’ use of the services will be governed by archived terms only until the end of the applicable Subscription Term.

Risk: After the current Subscription Term ends, renewal may move the customer to newer terms, potentially without a separate negotiation. Customers should review the renewal terms and any updated policies before renewal.

3. Stronger “clickwrap” and exclusivity language

The agreement now expressly states that clicking accept, accessing, using, or otherwise indicating acceptance binds the customer to the agreement “to the exclusion of all other terms.”

Risk: Purchase orders, vendor terms, or other customer documents may be expressly rejected unless separately accepted by LaunchDarkly. This could affect negotiated protections, security commitments, or procurement terms.

4. Data protection provisions

The agreement continues to define Customer Data broadly as electronic data submitted by or for the customer. It also states that processing of personal data is governed by the Data Processing Addendum (DPA).

LaunchDarkly may use or disclose anonymized or aggregated data for business purposes, including industry analysis, benchmarking, and analytics.

Risk: The agreement does not clearly explain the anonymization standard, whether data can be reidentified, or how long such data may be retained. Customers should review the DPA and privacy policy for details.

5. AI-model training

No express change concerning training AI models is apparent in the supplied diff. The revised text does not appear to add or remove language expressly permitting LaunchDarkly to use Customer Data, prompts, outputs, telemetry, or other customer information to train artificial-intelligence or machine-learning models.

However, the broad business-purpose and anonymized/aggregated-data rights could potentially encompass data used for analytics or model development, depending on the definitions and other incorporated policies.

Recommended action: Confirm the current Privacy Policy, DPA, product-specific terms, and any AI policy separately for an explicit prohibition or permission regarding AI training.

6. Other notable operational risks

• LaunchDarkly retains suspension rights, including suspension after notice for specified breaches.
• Free trials and free services remain largely “as-is,” with limited warranties and potentially broad customer liability.
• Automatic renewal and then-current pricing provisions remain important.
• LaunchDarkly receives authorization to use the customer’s name and logo in marketing materials.

Summary

Key change

• The diff states only that approximately 52 words were added, but does not provide the actual added language.

Legal and commercial risks

• The substantive effect of the changes cannot be assessed without seeing the new wording.
• It is not possible to determine whether the additions change:
• Liability or indemnification obligations
• Confidentiality requirements
• Data ownership or licensing rights
• Privacy or security commitments
• Termination rights
• Compliance obligations
• Use of subcontractors or service providers

AI training and customer data

• The provided diff contains no substantive language addressing AI, machine learning, model training, model improvement, or use of customer data.
• Accordingly, it is not possible to determine whether the contract:
• Permits customer data to be used to train or improve AI models
• Permits use of data after account termination
• Allows aggregated, de-identified, or anonymized data to be used for training
• Gives the provider ownership or a broad license over customer data
• Requires customer consent or provides an opt-out
• Protects confidential, personal, regulated, or proprietary information from AI training

Conclusion

The diff is insufficient for legal analysis. The approximately 52 added words—or a redline showing the actual additions—are needed to identify any new rights, obligations, or risks, particularly concerning the use of customer data to train AI models.

Structured Summary of Important Changes

1. Agreement structure and applicable terms

• The opening provisions were substantially reorganized. The prior reference to applicable terms being found in the archived policies page until the end of the applicable Subscription Term appears to have been replaced with a consolidated list of current policies and product-specific documents.
• The listed documents now include, among others:
• Terms of Service – Developer Tier and Foundation Self-Service
• Subscription Terms – Foundation (Invoiced) and Enterprise
• Professional Services Terms
• Security Program Addendum
• Privacy Policy
• Data Processing Addendum
• Subprocessors, Code of Conduct, Support Terms, SLAs, Acceptable Use Policy, and other policies.
• Risk: Customers may now be governed by multiple incorporated policies, some of which may be located on separate webpages and potentially updated independently. Customers should confirm which documents apply to their service and whether those documents can be changed unilaterally.

2. Contracting entity and acceptance

• The agreement now identifies the provider as Catamorphic, Co., a Delaware corporation, doing business as LaunchDarkly.
• The acceptance language has been revised to state that clicking to accept, accessing, using, or otherwise indicating acceptance of the service binds the customer to the agreement “to the exclusion of all other terms.”
• Risk: This strengthens the “exclusive terms” position and may reject conflicting terms in purchase orders, procurement portals, or customer forms unless separately accepted by LaunchDarkly.

3. Customer Data and data-use provisions

• The definition of “Customer Data” remains broadly framed as electronic data or information submitted by or on behalf of the customer, including through third-party services.
• The agreement continues to state that LaunchDarkly may collect, use, and disclose anonymized or aggregated data for business purposes, including industry analysis, benchmarking, and analytics.
• A related confidentiality provision continues to permit compilation and use or disclosure of aggregate data concerning the customer’s use of the service.
• AI training: The provided diff does not expressly add, remove, or modify language authorizing LaunchDarkly to use Customer Data to train artificial-intelligence or machine-learning models. There is no explicit AI-training permission or prohibition shown.
• Risk: Because the agreement permits broad use of anonymized or aggregated data, customers should verify whether AI-related usage could fall within “analytics,” benchmarking, or business purposes, and whether the DPA or Privacy Policy separately addresses model training, de-identification standards, retention, or re-identification.

4. Other notable changes

• Numerous edits standardize punctuation, capitalization, quotation marks, and apostrophes without apparent substantive effect.
• The agreement continues to incorporate the Acceptable Use Policy, DPA, product-specific terms, support terms, and other online policies.
• LaunchDarkly retains suspension rights for specified risks, including suspected fraud, unlawful activity, or breach, with a 15-day notice provision for certain breaches.
• LaunchDarkly may continue using the customer’s name and logo in marketing materials to identify the customer as a LaunchDarkly customer.

Summary

The diff only states that approximately 52 words were added; it does not include the actual added language.

Legal and Commercial Impact

• The specific changes cannot be analyzed without seeing the text of the additions.
• It is not possible to determine whether the additions:
• Expand or restrict either party’s rights;
• Create new obligations, warranties, or liabilities;
• Change confidentiality, security, ownership, or data-processing terms;
• Modify termination, indemnification, audit, or compliance provisions; or
• Introduce new fees, usage limits, or service conditions.

Customer Data and AI Training

No conclusion can be reached about AI-model training from the supplied diff. The missing language may potentially address whether customer data, prompts, outputs, usage information, or other content may be:

• Used to train, fine-tune, evaluate, or improve AI models;
• Shared with affiliates, vendors, or third-party model providers;
• Retained after termination or deletion;
• Used in aggregated, de-identified, or anonymized form; or
• Excluded from training by default or only excluded if the customer opts out.

These provisions can materially affect confidentiality, intellectual-property ownership, privacy compliance, data retention, and the customer’s control over its information.

Information Needed

Please provide the actual 52 added words, using the stated {addition} notation, so the legal changes and any AI-training risks can be assessed accurately.