24/7 support for active customersSales Mon–Fri, 8am–5pm
All Systems Operational(315) 754-HOST
Talk to usCustomer WebmailCustomer Login
Spark Rack Legal Center

Artificial Intelligence Policy

1. Purpose and Scope

This Artificial Intelligence Policy establishes Spark Rack’s rules, commitments, restrictions, and operating principles concerning artificial intelligence, machine learning, automated decision-making, model training, data licensing, consumer information, Customer Data, Customer Content, and related technologies.

This Policy applies to:

  1. Spark Rack;
  2. Grand Bay Collective, to the extent it operates or supports Spark Rack;
  3. The Spark Rack Network;
  4. All Spark Rack products and Services;
  5. All Spark Rack websites and customer-facing systems;
  6. The Spark Rack Customer Portal;
  7. All Spark Rack employees, contractors, representatives, and authorized personnel;
  8. All Service Providers processing information on Spark Rack’s behalf;
  9. All partners, vendors, subprocessors, and infrastructure providers receiving Spark Rack information;
  10. All Personal Information processed by Spark Rack;
  11. All Customer Data and Customer Content processed through the Services; and
  12. Any artificial-intelligence, machine-learning, model-based, generative, predictive, or automated technology that could interact with Spark Rack information.

This Policy is incorporated into and forms part of the Spark Rack Terms of Service and Privacy Policy.

2. Core Commitment

Spark Rack does not sell, lease, rent, license, trade, transmit, disclose, contribute, provide, exchange, or otherwise make consumer information, Personal Information, Customer Data, or Customer Content available for artificial-intelligence training, machine-learning training, model development, model improvement, model evaluation, data labeling, or any substantially similar purpose.

Spark Rack does not presently engage in these practices, has never adopted these practices as part of its Services, and has no present or planned business model involving the commercialization, licensing, transfer, or exploitation of customer or consumer information for artificial-intelligence or model-training purposes.

Spark Rack does not authorize any third party, vendor, contractor, partner, Service Provider, subprocessor, affiliate, infrastructure provider, software provider, data broker, advertising company, analytics company, or artificial-intelligence developer to use Spark Rack information for artificial-intelligence or machine-learning purposes.

3. Definitions

3.1 “Artificial Intelligence” or “AI”

Any technology, system, software, model, service, or process that performs or assists with functions commonly associated with artificial intelligence, machine learning, generative artificial intelligence, neural networks, deep learning, large language models, foundation models, computer vision, speech recognition, automated reasoning, predictive analytics, autonomous decision-making, or similar model-based processing.

3.2 “AI Training”

Any use of information to create, train, pretrain, fine-tune, retrain, adapt, align, reinforce, benchmark, test, evaluate, validate, improve, optimize, operate, or maintain an AI system or model.

AI Training includes:

  1. Pretraining;
  2. Supervised training;
  3. Unsupervised training;
  4. Semi-supervised training;
  5. Reinforcement learning;
  6. Reinforcement learning from human feedback;
  7. Fine-tuning;
  8. Prompt tuning;
  9. Instruction tuning;
  10. Model distillation;
  11. Transfer learning;
  12. Embedding generation for model development;
  13. Retrieval-augmented generation dataset creation;
  14. Model evaluation;
  15. Model benchmarking;
  16. Safety evaluation;
  17. Red-team evaluation;
  18. Quality scoring;
  19. Human review for model improvement;
  20. Data annotation;
  21. Data labeling;
  22. Creation of synthetic training data;
  23. Creation of preference datasets;
  24. Model monitoring intended to improve model performance;
  25. Prompt retention for product improvement;
  26. Response retention for model improvement;
  27. Model personalization based on consumer information;
  28. Creation of behavioral profiles for model use; and
  29. Any substantially similar activity.

3.3 “Automated Decision-Making”

The use of an automated system to make, recommend, materially influence, or substantially assist with a decision concerning a person, Account, Customer, End User, transaction, Service, payment, support matter, security matter, legal matter, or contractual relationship.

3.4 “Consumer Information”

Any information that identifies, relates to, describes, concerns, is associated with, or could reasonably be linked to an individual, household, Customer, prospective Customer, End User, account holder, authorized user, website visitor, support contact, or other person.

3.5 “Customer Content”

All files, websites, applications, databases, messages, emails, media, backups, credentials, configurations, records, logs, communications, documents, and other information submitted to, stored on, transmitted through, or processed by the Services at Customer’s direction.

3.6 “Customer Data”

Customer Content and all other information submitted to, collected through, generated by, or associated with Customer’s use of the Services.

3.7 “Model”

Any computational, statistical, algorithmic, predictive, generative, classification, recommendation, language, vision, speech, behavioral, or decision system that is trained, calibrated, adapted, or improved using data.

3.8 “Personal Information”

Information that identifies, relates to, describes, is reasonably capable of being associated with, or could reasonably be linked, directly or indirectly, with an identified or identifiable individual or household.

3.9 “Service Provider”

Any contractor, vendor, processor, subprocessor, partner, infrastructure provider, professional adviser, payment provider, domain provider, data-center provider, network provider, software provider, support provider, or other third party processing information on behalf of Spark Rack.

4. No AI Training of Any Kind

Spark Rack does not use Consumer Information, Personal Information, Customer Data, or Customer Content for AI Training.

This prohibition applies regardless of whether the information is:

  1. Identifiable;
  2. Pseudonymized;
  3. De-identified;
  4. Anonymized;
  5. Aggregated;
  6. Tokenized;
  7. Encrypted;
  8. Hashed;
  9. Redacted;
  10. Structured;
  11. Unstructured;
  12. Publicly available;
  13. Privately submitted;
  14. Generated automatically;
  15. Derived from service activity;
  16. Contained in logs;
  17. Contained in support communications;
  18. Contained in backups;
  19. Contained in network records;
  20. Contained in billing records;
  21. Contained in domain-registration records;
  22. Contained in email or messaging systems;
  23. Contained in website activity records; or
  24. Otherwise available to Spark Rack.

Spark Rack does not treat de-identification, aggregation, redaction, hashing, tokenization, or pseudonymization as permission to use information for AI Training.

5. No Sale, Lease, Rental, or Licensing for AI

Spark Rack does not and will not, under this Policy, sell, lease, rent, license, sublicense, trade, exchange, monetize, contribute, or commercially exploit Consumer Information, Personal Information, Customer Data, or Customer Content for AI Training or model-related purposes.

This prohibition includes:

  1. Direct sale of datasets;
  2. Licensing information to an AI developer;
  3. Providing information in exchange for credits, discounts, services, infrastructure, or other consideration;
  4. Leasing access to databases or storage;
  5. Providing model developers with access to Customer systems;
  6. Providing data feeds for AI development;
  7. Providing historical support records for model training;
  8. Providing website or Customer Portal activity for model development;
  9. Providing email or message content for language-model training;
  10. Providing network records for behavioral-model training;
  11. Providing billing or transaction records for predictive-model development;
  12. Providing Customer Content for generative-model development;
  13. Providing de-identified or aggregated information for model training;
  14. Providing information for data-labeling projects;
  15. Providing information for synthetic-data generation;
  16. Providing prompts or responses for AI improvement;
  17. Providing information to data brokers serving AI companies;
  18. Allowing third parties to scrape Spark Rack systems for AI datasets;
  19. Allowing a third party to create embeddings from Spark Rack information;
  20. Allowing a third party to retain Spark Rack information for future model use; and
  21. Any substantially similar arrangement.

6. No Transmission or Disclosure for AI Training

Spark Rack does not transmit, disclose, upload, submit, sync, copy, export, transfer, publish, or otherwise provide Consumer Information, Personal Information, Customer Data, or Customer Content to an AI service for training, model improvement, model evaluation, or unrelated AI processing.

Spark Rack personnel may not:

  1. Paste Customer information into a public AI chatbot;
  2. Paste Customer information into a private AI chatbot;
  3. Upload support tickets to an AI service;
  4. Upload Customer files to an AI service;
  5. Upload server logs to an AI service;
  6. Upload network records to an AI service;
  7. Upload emails or messages to an AI service;
  8. Upload screenshots containing Customer information to an AI service;
  9. Upload invoices or billing records to an AI service;
  10. Upload domain-registration records to an AI service;
  11. Use an AI transcription service for Customer communications;
  12. Use an AI summarization service for support requests;
  13. Use an AI coding assistant with Customer source code, credentials, configurations, or proprietary information;
  14. Use an AI image service with Customer images or media;
  15. Use an AI voice service with Customer recordings;
  16. Use AI tools to review legal, abuse, security, or privacy requests containing Personal Information;
  17. Use AI tools to generate responses based on private Customer information;
  18. Use AI browser extensions capable of reading Customer information;
  19. Enable AI features in software that would transmit Customer information to a model provider; or
  20. Otherwise expose Spark Rack information to AI processing without an express policy exception approved in writing.

7. No Third-Party Authorization

No Service Provider, contractor, partner, vendor, affiliate, subprocessor, infrastructure provider, or other third party is authorized by Spark Rack to use Consumer Information, Personal Information, Customer Data, or Customer Content for AI Training.

Third parties processing information on Spark Rack’s behalf are prohibited from:

  1. Training a model using Spark Rack information;
  2. Improving a model using Spark Rack information;
  3. Testing a model using Spark Rack information;
  4. Evaluating a model using Spark Rack information;
  5. Benchmarking a model using Spark Rack information;
  6. Creating embeddings from Spark Rack information for unrelated purposes;
  7. Creating training datasets from Spark Rack information;
  8. Creating synthetic datasets based on Spark Rack information;
  9. Retaining prompts containing Spark Rack information;
  10. Retaining model responses derived from Spark Rack information;
  11. Performing human review for model improvement;
  12. Annotating Spark Rack information for AI development;
  13. Combining Spark Rack information with external datasets for model use;
  14. Using Spark Rack information to create profiles or predictions;
  15. Selling Spark Rack information to an AI developer;
  16. Licensing Spark Rack information to an AI developer;
  17. Disclosing Spark Rack information to a subprocessor for AI use;
  18. Using Spark Rack information to improve a generally available product;
  19. Using Spark Rack information for research involving AI models; or
  20. Using Spark Rack information for any purpose beyond the limited service authorized by Spark Rack.

8. Contractual Restrictions on Service Providers

Where reasonably appropriate, Spark Rack requires Service Providers to agree that they will:

  1. Process information only for defined and authorized purposes;
  2. Follow Spark Rack’s documented instructions;
  3. Maintain confidentiality;
  4. Use appropriate security safeguards;
  5. Limit access to authorized personnel;
  6. Not sell Personal Information;
  7. Not rent Personal Information;
  8. Not share Personal Information for targeted advertising;
  9. Not retain information beyond the authorized purpose;
  10. Not combine information for unauthorized purposes;
  11. Not use information for AI Training;
  12. Not use information for machine-learning development;
  13. Not use information for model improvement;
  14. Not use information for model evaluation;
  15. Not use information for profiling;
  16. Not create embeddings for unrelated purposes;
  17. Not use information to create synthetic data;
  18. Not disclose information to an AI provider;
  19. Not enable optional AI features that process Spark Rack information;
  20. Apply equivalent restrictions to authorized subprocessors;
  21. Notify Spark Rack of any unauthorized AI use;
  22. Delete or return information when no longer needed;
  23. Cooperate with investigations and audits concerning misuse; and
  24. Comply with Applicable Law and this Policy.

9. No AI in Spark Rack Products or Services

Spark Rack does not use artificial intelligence, generative artificial intelligence, large language models, machine-learning models, neural networks, or similar model-based systems as part of Spark Rack’s products or Services.

Spark Rack does not use AI to operate or provide:

  1. The Customer Portal;
  2. Account registration;
  3. Account authentication;
  4. Account management;
  5. Billing;
  6. Payment processing decisions;
  7. Invoice generation;
  8. Hosting services;
  9. Server services;
  10. Network services;
  11. The Spark Rack Network;
  12. Domain registration;
  13. DNS services;
  14. Email services;
  15. Messaging services;
  16. Backup services;
  17. Storage services;
  18. Monitoring services;
  19. Managed services;
  20. Customer support;
  21. Security administration;
  22. Fraud determinations;
  23. Abuse determinations;
  24. Content review;
  25. Legal review;
  26. Privacy-request review;
  27. Employment decisions;
  28. Customer communications;
  29. Service recommendations;
  30. Pricing decisions;
  31. Service suspension decisions;
  32. Service termination decisions; or
  33. Any other Spark Rack product or Service.

10. No AI-Based Customer Support

Spark Rack does not use AI to:

  1. Read support tickets;
  2. Summarize support tickets;
  3. Classify support tickets;
  4. Assign support priority;
  5. Generate support responses;
  6. Recommend technical actions;
  7. Review uploaded support files;
  8. Analyze Customer logs;
  9. Analyze Customer screenshots;
  10. Transcribe support calls;
  11. Score Customer sentiment;
  12. Assess Customer behavior;
  13. Identify Customer emotions;
  14. Predict Customer satisfaction;
  15. Evaluate support personnel;
  16. Create support knowledge from private Customer communications;
  17. Train support chatbots;
  18. Operate an automated support chatbot; or
  19. Replace human support review with model-based processing.

Support requests are handled through human review, traditional software workflows, and predefined administrative processes.

11. No AI-Based Security or Abuse Decisions

Spark Rack does not use AI or machine-learning models to make final decisions concerning:

  1. Account compromise;
  2. Fraud;
  3. Spam;
  4. Phishing;
  5. Malware;
  6. Network Abuse;
  7. Policy violations;
  8. Service suspension;
  9. Service termination;
  10. Customer eligibility;
  11. Payment acceptance;
  12. Chargeback disputes;
  13. Identity verification;
  14. Sanctions compliance;
  15. Legal compliance;
  16. Privacy requests;
  17. Employment decisions; or
  18. Any decision that could materially affect a Customer or consumer.

Spark Rack may use deterministic, rule-based controls, known signatures, fixed thresholds, blocklists, allowlists, firewall rules, rate limits, and human review to protect the Services.

12. No Automated Profiling

Spark Rack does not use AI to create or infer profiles concerning:

  1. Customer behavior;
  2. Consumer interests;
  3. Purchasing preferences;
  4. Financial condition;
  5. Creditworthiness;
  6. Personality;
  7. Emotions;
  8. Health;
  9. Political beliefs;
  10. Religious beliefs;
  11. Sexual orientation;
  12. Race or ethnicity;
  13. Citizenship or immigration status;
  14. Employment suitability;
  15. Likelihood of cancellation;
  16. Likelihood of purchasing a Service;
  17. Likelihood of committing fraud;
  18. Likelihood of violating a policy;
  19. Support priority;
  20. Customer value;
  21. Advertising eligibility;
  22. Location patterns;
  23. Communication patterns;
  24. Website behavior;
  25. Network behavior;
  26. Risk classifications; or
  27. Any other personal characteristic or prediction.

13. No AI-Based Advertising or Marketing

Spark Rack does not use AI or machine learning to:

  1. Create advertising profiles;
  2. Target advertisements;
  3. Personalize advertising;
  4. Predict purchasing behavior;
  5. Score marketing leads;
  6. Generate individualized pricing;
  7. Generate promotional messages from private Customer information;
  8. Analyze private Customer communications for marketing purposes;
  9. Create lookalike audiences;
  10. Combine information from unrelated services;
  11. Perform cross-context behavioral advertising;
  12. Sell audience segments;
  13. License behavioral data;
  14. Optimize advertising using Customer Data; or
  15. Develop marketing models using consumer information.

14. No AI-Based Pricing or Eligibility Decisions

Spark Rack does not use AI to determine:

  1. Service pricing;
  2. Discount eligibility;
  3. Promotional eligibility;
  4. Payment terms;
  5. Credit terms;
  6. Refund eligibility;
  7. Account approval;
  8. Service availability;
  9. Service restrictions;
  10. Security deposits;
  11. Support access;
  12. Support priority;
  13. Service suspension;
  14. Service termination;
  15. Contract terms; or
  16. Any other material condition of service.

15. No Facial, Voice, Emotion, or Biometric AI

Spark Rack does not use Customer or consumer information for:

  1. Facial recognition;
  2. Face matching;
  3. Facial analysis;
  4. Voice recognition;
  5. Voiceprint creation;
  6. Speaker identification;
  7. Voice cloning;
  8. Emotion recognition;
  9. Sentiment scoring through AI;
  10. Gait recognition;
  11. Behavioral biometrics;
  12. Keystroke profiling;
  13. Biometric categorization;
  14. Biometric model training;
  15. Image-model training;
  16. Speech-model training;
  17. Audio-model training; or
  18. Any substantially similar biometric or inferential use.

16. No AI Use of Support, Billing, or Account Records

The following records are not used for AI Training or model development:

  1. Support tickets;
  2. Support emails;
  3. Support attachments;
  4. Support screenshots;
  5. Support logs;
  6. Telephone notes;
  7. Voicemail messages;
  8. Chat messages;
  9. Invoices;
  10. Payment records;
  11. Refund records;
  12. Account credits;
  13. Chargeback records;
  14. Tax records;
  15. Identity-verification records;
  16. Account contact information;
  17. Account login information;
  18. Account preferences;
  19. Service selections;
  20. Service configurations;
  21. Cancellation requests;
  22. Customer complaints;
  23. Survey responses;
  24. Legal notices;
  25. Privacy requests;
  26. Abuse reports;
  27. Security reports; or
  28. Any other administrative record.

17. No AI Use of Hosting, Network, Email, or DNS Data

The following technical information is not used for AI Training or model development:

  1. Website files;
  2. Application files;
  3. Source code;
  4. Databases;
  5. Database queries;
  6. Database contents;
  7. Backups;
  8. Storage objects;
  9. Email content;
  10. Email metadata;
  11. Mailbox information;
  12. Message headers;
  13. DNS records;
  14. DNS metadata;
  15. Domain-registration records;
  16. Network-flow information;
  17. Source and destination IP addresses;
  18. Source and destination ports;
  19. Traffic volume;
  20. Connection timestamps;
  21. Routing information;
  22. Firewall records;
  23. Authentication records;
  24. Login history;
  25. Security-event records;
  26. Server logs;
  27. Application logs;
  28. Resource-usage records;
  29. Monitoring records;
  30. Backup-job records;
  31. DDoS-mitigation records;
  32. Packet captures;
  33. Diagnostic records;
  34. Configuration records;
  35. API requests;
  36. API responses;
  37. Cookies;
  38. Session information;
  39. Browser information;
  40. Device information; or
  41. Any other technical or operational information.

18. No AI Use of Aggregated or De-Identified Information

Spark Rack does not use aggregated, statistical, pseudonymized, anonymized, or de-identified Customer or consumer information for AI Training.

Spark Rack may use limited aggregated or de-identified information for traditional, non-AI purposes such as:

  1. Capacity planning;
  2. Availability reporting;
  3. Financial reporting;
  4. Resource measurement;
  5. Service planning;
  6. Security reporting;
  7. Incident review;
  8. Traditional statistical analysis;
  9. Human quality review;
  10. Performance measurement; and
  11. Internal business administration.

Aggregation or de-identification does not create an exception to Spark Rack’s prohibition on AI Training.

19. Permitted Non-AI Automation

Spark Rack may use traditional, deterministic, rule-based software to perform routine operational functions.

Permitted non-AI automation may include:

  1. Generating invoices;
  2. Processing scheduled renewals;
  3. Sending service notices;
  4. Sending payment notices;
  5. Provisioning Services;
  6. Executing backups;
  7. Applying firewall rules;
  8. Applying rate limits;
  9. Blocking known malicious signatures;
  10. Detecting repeated failed logins using fixed thresholds;
  11. Monitoring service availability;
  12. Measuring resource usage;
  13. Rotating logs;
  14. Scheduling maintenance;
  15. Processing domain-renewal schedules;
  16. Routing support requests based on user-selected categories;
  17. Applying predefined email-filtering rules;
  18. Applying predefined network controls;
  19. Generating status notifications;
  20. Managing session expiration;
  21. Processing user-configured alerts;
  22. Running predefined health checks; and
  23. Performing other fixed, non-learning operational tasks.

Permitted rule-based automation does not:

  1. Train on Customer information;
  2. Learn from Customer behavior;
  3. Create a model;
  4. Generate content;
  5. Infer personal characteristics;
  6. Create behavioral profiles;
  7. Make model-based predictions;
  8. Continuously adapt based on consumer information; or
  9. Use Customer information for unrelated purposes.

20. Customer-Controlled AI Workloads

A Customer may independently choose to install, host, operate, or connect third-party AI software using Customer-controlled infrastructure, provided the activity complies with the Terms of Service, Acceptable Use Policy, Applicable Law, resource limits, and all service-specific requirements.

Customer-controlled AI activity:

  1. Is not a Spark Rack product;
  2. Is not operated by Spark Rack;
  3. Is not endorsed by Spark Rack;
  4. Is controlled by Customer;
  5. Is subject to Customer’s own privacy obligations;
  6. Is subject to Customer’s own contractual obligations;
  7. Must not expose Spark Rack information without authorization;
  8. Must not use another Customer’s information;
  9. Must not create unlawful or prohibited content;
  10. Must not interfere with the Spark Rack Network;
  11. Must not evade resource limits;
  12. Must not create an unreasonable security risk;
  13. Must not use Spark Rack personnel or systems as an AI data source;
  14. Must not imply that Spark Rack sponsors or operates the AI system; and
  15. Does not change Spark Rack’s prohibition on AI use of Customer or consumer information.

21. Customer-Selected Third-Party AI Integrations

A Customer may independently configure a Customer-controlled application or Service to transmit Customer Data to a third-party AI provider.

Where Customer independently selects, configures, authorizes, or instructs such an integration:

  1. Customer is responsible for the integration;
  2. Customer is responsible for reviewing the third party’s privacy policy;
  3. Customer is responsible for reviewing the third party’s training and retention terms;
  4. Customer is responsible for obtaining required consent;
  5. Customer is responsible for providing required notices;
  6. Customer is responsible for protecting credentials;
  7. Customer is responsible for limiting the information transmitted;
  8. Customer is responsible for honoring privacy rights;
  9. Customer is responsible for compliance with Applicable Law;
  10. Customer is responsible for the third party’s use of Customer Data;
  11. The third party is not acting on Spark Rack’s behalf merely because the integration uses Spark Rack infrastructure;
  12. Spark Rack does not authorize the third party to use Spark Rack information;
  13. Spark Rack does not assume responsibility for the third party’s privacy practices; and
  14. Spark Rack may restrict the integration if it violates the Terms of Service or Acceptable Use Policy.

22. Personnel Restrictions

Spark Rack personnel, contractors, and representatives must not:

  1. Submit Personal Information to an AI service;
  2. Submit Customer Data to an AI service;
  3. Submit Customer Content to an AI service;
  4. Submit support tickets to an AI service;
  5. Submit Customer source code to an AI service;
  6. Submit private configurations to an AI service;
  7. Submit credentials to an AI service;
  8. Submit private keys or tokens to an AI service;
  9. Submit Customer logs to an AI service;
  10. Submit private network information to an AI service;
  11. Use AI to generate Customer-facing support answers based on private information;
  12. Use AI to review legal or privacy requests;
  13. Use AI to review abuse or security reports;
  14. Use AI to determine whether to suspend or terminate a Customer;
  15. Use AI to score or profile Customers;
  16. Use AI to create marketing profiles;
  17. Use AI transcription on private Customer calls;
  18. Enable unapproved AI features in business software;
  19. Use browser-integrated AI tools on pages displaying Customer information;
  20. Use AI tools to summarize private communications;
  21. Use AI tools to generate legal or contractual decisions;
  22. Use AI tools to process Sensitive Personal Information;
  23. Use Customer information to test an AI product;
  24. Use Customer information for personal AI projects; or
  25. Otherwise expose Customer or consumer information to AI processing.

23. Software and Vendor Review

Before adopting software or a Service Provider that may include AI functionality, Spark Rack may evaluate:

  1. Whether AI features are present;
  2. Whether AI features are enabled by default;
  3. Whether information is transmitted to an AI provider;
  4. Whether prompts or responses are retained;
  5. Whether information is used for model training;
  6. Whether model training can be disabled;
  7. Whether AI processing can be disabled entirely;
  8. Whether the provider uses subprocessors;
  9. Whether browser extensions or integrations can read Customer information;
  10. Whether telemetry includes Customer information;
  11. Whether data is combined with external information;
  12. Whether the provider claims rights to submitted information;
  13. Whether information is retained after termination;
  14. Whether the provider offers appropriate contractual restrictions;
  15. Whether the provider permits auditing or verification;
  16. Whether the provider can comply with deletion requests;
  17. Whether the provider can isolate Spark Rack information;
  18. Whether the provider can prohibit model training contractually;
  19. Whether the provider’s practices are consistent with this Policy; and
  20. Whether use of the provider creates an unacceptable privacy or security risk.

Spark Rack may reject or discontinue a product or Service Provider whose AI practices are inconsistent with this Policy.

24. AI Features Embedded in Third-Party Products

If a third-party product used by Spark Rack introduces an AI feature, Spark Rack may:

  1. Disable the feature;
  2. Restrict access to the feature;
  3. Block the feature from receiving Customer information;
  4. Modify the product configuration;
  5. Remove the integration;
  6. Replace the product;
  7. Require contractual assurances;
  8. Request deletion of transmitted information;
  9. Investigate whether unauthorized processing occurred;
  10. Notify affected Customers where appropriate or legally required;
  11. Suspend information flows to the provider;
  12. Terminate the provider relationship; or
  13. Take another reasonable protective action.

The mere presence of an optional AI feature in third-party software does not authorize its use.

25. No Data Brokerage for AI

Spark Rack does not provide Customer or consumer information to:

  1. Data brokers;
  2. Dataset marketplaces;
  3. AI data exchanges;
  4. Model developers;
  5. Foundation-model providers;
  6. Generative-AI providers;
  7. Data-labeling companies;
  8. Synthetic-data companies;
  9. Advertising-data providers;
  10. Behavioral-profiling companies;
  11. Facial-recognition companies;
  12. Voice-recognition companies;
  13. Biometric-model developers;
  14. AI research organizations;
  15. AI benchmarking services;
  16. AI safety-testing organizations;
  17. AI prompt marketplaces;
  18. AI plugin providers;
  19. AI browser-extension providers; or
  20. Any other entity seeking information for model-related use.

26. No AI Research Use

Spark Rack does not use or provide Customer or consumer information for academic, commercial, nonprofit, governmental, internal, or external AI research.

This prohibition includes research involving:

  1. Language models;
  2. Computer vision;
  3. Speech recognition;
  4. Voice synthesis;
  5. Facial recognition;
  6. Behavioral prediction;
  7. Security models;
  8. Fraud models;
  9. Recommendation systems;
  10. Advertising models;
  11. Biometric systems;
  12. Emotion recognition;
  13. Medical models;
  14. Financial models;
  15. Autonomous agents;
  16. Generative systems;
  17. Foundation models;
  18. Dataset development;
  19. Model safety research; or
  20. Any substantially similar area.

27. No Prompt, Response, or Embedding Retention for AI

Spark Rack does not create, retain, license, disclose, or supply:

  1. AI prompts containing Customer information;
  2. AI responses derived from Customer information;
  3. Prompt-and-response pairs;
  4. Conversation datasets;
  5. Instruction datasets;
  6. Preference datasets;
  7. Embedding databases for AI development;
  8. Vector representations for unrelated model use;
  9. Model feedback records;
  10. Human-rating datasets;
  11. Model-evaluation datasets;
  12. Red-team datasets;
  13. Synthetic datasets based on Customer information; or
  14. Any substantially similar model-development material.

28. No AI-Based Content Generation

Spark Rack does not use AI to generate:

  1. Customer support responses;
  2. Legal notices;
  3. Privacy responses;
  4. Abuse determinations;
  5. Security determinations;
  6. Billing decisions;
  7. Invoices;
  8. Contract decisions;
  9. Customer-specific recommendations;
  10. Customer profiles;
  11. Service suspension notices;
  12. Service termination notices;
  13. Incident reports based on private Customer information;
  14. Marketing content based on Customer behavior;
  15. Personalized advertisements;
  16. Employment decisions;
  17. Performance evaluations based on private communications; or
  18. Any other output derived from private Customer or consumer information.

29. No AI-Based Legal or Contractual Interpretation

Spark Rack does not use AI as the final authority to:

  1. Interpret contracts;
  2. Interpret Customer obligations;
  3. Interpret privacy rights;
  4. Interpret legal process;
  5. Determine compliance with law;
  6. Determine whether a subpoena is valid;
  7. Determine whether to disclose information;
  8. Determine whether an Account violated a policy;
  9. Determine whether to suspend or terminate Services;
  10. Resolve billing disputes;
  11. Resolve privacy disputes;
  12. Resolve copyright complaints;
  13. Resolve trademark complaints;
  14. Resolve abuse complaints;
  15. Respond to law enforcement;
  16. Determine sanctions status;
  17. Determine export-control status; or
  18. Make another legally significant determination.

30. No Hidden or Secondary AI Use

Spark Rack prohibits secondary or undisclosed use of information for AI purposes.

Information collected for billing, support, security, networking, hosting, domain registration, email delivery, backups, legal compliance, or another operational purpose may not later be repurposed for:

  1. Model training;
  2. Model testing;
  3. Model improvement;
  4. Profiling;
  5. Prediction;
  6. Recommendation systems;
  7. Behavioral analysis;
  8. Advertising optimization;
  9. Data brokerage;
  10. AI research;
  11. Data labeling;
  12. Embedding generation;
  13. Synthetic-data creation;
  14. Prompt libraries;
  15. Human-feedback datasets; or
  16. Any substantially similar purpose.

31. Data Minimization

Spark Rack limits collection and processing of information to what is reasonably necessary to:

  1. Provide the Services;
  2. Maintain Accounts;
  3. Process billing and payments;
  4. Provide support;
  5. Maintain security;
  6. Prevent fraud and Abuse;
  7. Operate the Spark Rack Network;
  8. Comply with law;
  9. Resolve disputes;
  10. Maintain business records;
  11. Protect legal rights;
  12. Communicate with Customers; and
  13. Operate Spark Rack’s business through non-AI methods.

Spark Rack does not collect additional information for the purpose of building AI datasets or training models.

32. Data Retention

Spark Rack retains information only for authorized operational, contractual, security, legal, financial, or administrative purposes.

Information is not retained for:

  1. Future AI Training;
  2. Future model development;
  3. Future model evaluation;
  4. Future profiling;
  5. Future advertising optimization;
  6. Creation of training datasets;
  7. Creation of prompt libraries;
  8. Creation of embedding databases;
  9. Creation of synthetic datasets;
  10. Sale or licensing to AI companies; or
  11. Any speculative model-related use.

33. Security of Information

Spark Rack uses reasonable safeguards designed to prevent unauthorized access, use, disclosure, transfer, or exploitation of information.

Safeguards may include:

  1. Access controls;
  2. Role-based permissions;
  3. Multifactor authentication;
  4. Unique credentials;
  5. Network segmentation;
  6. Encryption where appropriate and supported;
  7. Firewall controls;
  8. Logging;
  9. Traditional security monitoring;
  10. Rule-based alerts;
  11. Vendor review;
  12. Contractual restrictions;
  13. Personnel confidentiality obligations;
  14. Data-minimization practices;
  15. Secure deletion;
  16. Incident-response procedures;
  17. Credential rotation;
  18. Administrative review;
  19. Restrictions on AI tools;
  20. Blocking or disabling unauthorized AI integrations; and
  21. Other safeguards appropriate to the Services.

34. Unauthorized AI Use as a Security Incident

Spark Rack may treat unauthorized use, disclosure, transmission, or retention of information by an AI system as a security or privacy incident.

Response actions may include:

  1. Disabling the affected integration;
  2. Blocking information transfers;
  3. Revoking credentials;
  4. Suspending access;
  5. Preserving relevant records;
  6. Investigating the scope of the event;
  7. Identifying affected information;
  8. Requesting deletion from the AI provider;
  9. Requesting written confirmation of deletion;
  10. Notifying affected Customers where appropriate or legally required;
  11. Notifying regulators where legally required;
  12. Terminating a Service Provider relationship;
  13. Disciplining personnel;
  14. Updating controls;
  15. Reviewing related software and integrations;
  16. Documenting corrective actions; and
  17. Taking any other reasonable remedial measure.

35. Reporting Suspected Unauthorized AI Use

A person who believes Consumer Information, Personal Information, Customer Data, or Customer Content has been used or disclosed for unauthorized AI purposes should report the matter through the appropriate privacy, legal, security, or support channel in the Spark Rack Customer Portal.

Written reports may be mailed to:

Spark Rack
Attn: AI Policy and Privacy
PO Box 2215
Valdosta, GA 31604
United States

A report should include, where available:

  1. The reporter’s name and contact information;
  2. The affected Account or Service;
  3. A description of the suspected AI use;
  4. The date or time period involved;
  5. The third party or system involved;
  6. The categories of information affected;
  7. Relevant screenshots, notices, or documentation;
  8. Any action already taken;
  9. The requested response; and
  10. Any other information reasonably necessary to investigate the report.

36. Investigation and Enforcement

Spark Rack may investigate suspected violations of this Policy.

Investigative or enforcement actions may include:

  1. Reviewing system configurations;
  2. Reviewing vendor settings;
  3. Reviewing access logs;
  4. Reviewing transmission records;
  5. Reviewing contractual terms;
  6. Requesting information from personnel or Service Providers;
  7. Disabling AI features;
  8. Removing software;
  9. Blocking integrations;
  10. Revoking access;
  11. Suspending a Service Provider;
  12. Terminating a Service Provider;
  13. Requiring deletion of information;
  14. Requiring written certification of deletion;
  15. Restricting personnel access;
  16. Applying disciplinary measures;
  17. Notifying affected Customers;
  18. Notifying authorities where required;
  19. Preserving evidence;
  20. Updating technical safeguards;
  21. Updating contractual safeguards; and
  22. Taking other reasonable corrective action.

37. Customer Rights and Requests

A Customer or consumer may contact Spark Rack to ask whether Spark Rack:

  1. Uses Personal Information for AI Training;
  2. Uses Customer Data for AI Training;
  3. Uses Customer Content for AI Training;
  4. Sells information for AI Training;
  5. Leases information for AI Training;
  6. Licenses information for AI Training;
  7. Discloses information to an AI provider;
  8. Uses AI for profiling;
  9. Uses AI for automated decisions;
  10. Uses AI in Customer support;
  11. Uses AI in security or fraud review;
  12. Uses AI in billing or pricing;
  13. Uses AI in marketing;
  14. Uses AI in employment decisions;
  15. Allows Service Providers to train models using Spark Rack information; or
  16. Has experienced a known unauthorized AI disclosure affecting the requester.

Subject to verification and Applicable Law, Spark Rack will respond to appropriate privacy requests through the procedures described in the Spark Rack Privacy Policy.

38. No AI Opt-Out Is Necessary

Spark Rack does not offer an AI-training opt-out because Spark Rack does not use Customer or consumer information for AI Training in the first instance.

No Customer or consumer is required to:

  1. Disable AI Training;
  2. Submit an opt-out request;
  3. Change an Account setting;
  4. Pay an additional fee;
  5. Select a privacy tier;
  6. Use a special Service plan;
  7. Submit a Global Privacy Control signal;
  8. Contact support;
  9. Negotiate a separate contract; or
  10. Take any other action

to prevent Spark Rack from using information for AI Training, because such use is prohibited by default.

39. No Reduced Service for Rejecting AI Use

Because Spark Rack does not use information for AI Training, Customers do not receive:

  1. Higher prices for rejecting AI use;
  2. Lower service quality for rejecting AI use;
  3. Reduced support for rejecting AI use;
  4. Reduced functionality for rejecting AI use;
  5. Lower resource limits for rejecting AI use;
  6. Reduced security for rejecting AI use;
  7. Fewer privacy protections for rejecting AI use; or
  8. Any other disadvantage connected to AI-data rights.

40. Policy for Future Technologies

This Policy applies to current and future technologies that perform functions substantially similar to artificial intelligence, machine learning, generative models, predictive models, neural networks, autonomous agents, or model-based automated decision systems, regardless of the terminology used by the technology provider.

A provider may not avoid this Policy by describing a technology as:

  1. Automation;
  2. Smart processing;
  3. Intelligent processing;
  4. Predictive processing;
  5. Advanced analytics;
  6. Cognitive computing;
  7. Adaptive software;
  8. Statistical learning;
  9. Decision intelligence;
  10. Content assistance;
  11. Copilot functionality;
  12. Generative search;
  13. Semantic processing;
  14. Automated insights;
  15. Enhanced assistance;
  16. Model-assisted processing;
  17. Agentic processing; or
  18. Another marketing or technical term.

41. Future Changes to Spark Rack’s Position

Spark Rack has no present plan to sell, lease, license, transmit, or otherwise provide Consumer Information, Personal Information, Customer Data, or Customer Content for AI Training.

Spark Rack has no present plan to introduce AI into Spark Rack products or Services.

Spark Rack has no present plan to authorize Service Providers or partners to use Spark Rack information for AI Training.

Any future proposal to materially reverse these commitments would require, at a minimum:

  1. A formal review of legal, privacy, security, contractual, and ethical implications;
  2. A written revision to this Policy;
  3. Clear and conspicuous advance notice;
  4. Identification of the categories of information involved;
  5. Identification of the specific purpose;
  6. Identification of the parties receiving the information;
  7. Identification of retention and deletion practices;
  8. Identification of the legal basis for processing;
  9. Any consent required by Applicable Law;
  10. Any contractual amendments required by Applicable Law or existing agreements;
  11. A meaningful ability to reject the new use where legally or contractually required;
  12. A prohibition on retroactive use of information collected under this Policy without legally sufficient authorization;
  13. Updated Service Provider restrictions;
  14. Updated security controls;
  15. Updated privacy-request procedures; and
  16. Any other safeguards required by Applicable Law.

No general update to the Terms of Service, Privacy Policy, or another policy will silently authorize retroactive AI Training of information collected while this Policy was in effect.

42. No Retroactive AI Training

Information collected, received, stored, or processed while this Policy is in effect will not be retroactively designated as AI-training data merely because Spark Rack later changes a product, vendor, policy, business model, or technical system.

Any legally valid future use would require a separate and specific legal basis and any notice, consent, or contractual agreement required under Applicable Law.

43. Conflicts with Other Documents

If another Spark Rack policy, Order, statement of work, product description, vendor agreement, integration, or communication appears to permit AI Training of Consumer Information, Personal Information, Customer Data, or Customer Content, this Policy controls unless a separately signed written agreement:

  1. Specifically identifies this Policy;
  2. Specifically identifies the provision being modified;
  3. Specifically identifies the information involved;
  4. Specifically identifies the permitted AI use;
  5. Specifically identifies the receiving party;
  6. Specifically identifies the retention period;
  7. Specifically identifies the applicable safeguards;
  8. Is signed by an authorized representative of Spark Rack;
  9. Is signed by the affected Customer where required;
  10. Complies with Applicable Law; and
  11. Obtains any legally required consumer consent.

No click-through vendor term, default software setting, hidden product feature, online service term, or unilateral third-party policy grants permission to use Spark Rack information for AI Training.

44. No Third-Party Rights

This Policy does not grant any AI provider, data broker, model developer, vendor, partner, researcher, or other third party a right to:

  1. Access Spark Rack systems;
  2. Scrape Spark Rack websites;
  3. Collect Customer information;
  4. Collect Customer Content;
  5. Use Spark Rack information for training;
  6. Use Spark Rack information for evaluation;
  7. Create embeddings from Spark Rack information;
  8. Retain Spark Rack information;
  9. Publish Spark Rack information;
  10. License Spark Rack information;
  11. Claim ownership of Spark Rack information;
  12. Use Spark Rack information for research; or
  13. Use Spark Rack information for any unauthorized purpose.

45. Severability

If any provision of this Policy is found unenforceable, invalid, or unlawful, that provision will be modified or severed only to the minimum extent necessary, and the remaining provisions will remain in effect.

The invalidity of one restriction does not create permission for AI Training, sale, leasing, licensing, transmission, disclosure, or model use beyond what Applicable Law requires.

46. Enforcement Rights

Spark Rack may enforce this Policy through:

  1. Contractual restrictions;
  2. Vendor termination;
  3. Access revocation;
  4. Technical blocking;
  5. Credential revocation;
  6. Data-deletion demands;
  7. Legal demands;
  8. Injunctive relief;
  9. Claims for damages;
  10. Indemnification rights;
  11. Regulatory complaints;
  12. Incident-response procedures;
  13. Personnel discipline;
  14. Service suspension;
  15. Service termination;
  16. Audit rights;
  17. Contractual remedies; and
  18. Any other remedy available under the Terms of Service, contract, or Applicable Law.

47. Contact Information

Questions, concerns, privacy requests, or reports concerning this Policy may be submitted through the appropriate privacy, legal, security, account, or support channel in the Spark Rack Customer Portal.

Written correspondence may be mailed to:

Spark Rack
Attn: AI Policy and Privacy
PO Box 2215
Valdosta, GA 31604
United States

48. Customer and Consumer Assurance

Spark Rack affirms the following without qualification:

  1. Spark Rack does not use Customer or consumer information for AI Training.
  2. Spark Rack does not sell Customer or consumer information for AI Training.
  3. Spark Rack does not lease Customer or consumer information for AI Training.
  4. Spark Rack does not rent Customer or consumer information for AI Training.
  5. Spark Rack does not license Customer or consumer information for AI Training.
  6. Spark Rack does not transmit Customer or consumer information for AI Training.
  7. Spark Rack does not disclose Customer or consumer information for AI Training.
  8. Spark Rack does not trade Customer or consumer information for AI Training.
  9. Spark Rack does not provide Customer or consumer information to data brokers for AI Training.
  10. Spark Rack does not provide Customer or consumer information to model developers.
  11. Spark Rack does not authorize Service Providers to train models using Spark Rack information.
  12. Spark Rack does not use AI to profile Customers or consumers.
  13. Spark Rack does not use AI to make significant decisions concerning Customers or consumers.
  14. Spark Rack does not use AI in Customer support.
  15. Spark Rack does not use AI in Spark Rack products or Services.
  16. Spark Rack has no present plan to adopt a business model based on selling, licensing, leasing, or transmitting Customer or consumer information for AI development.
  17. Spark Rack does not require an AI opt-out because AI Training is prohibited by default.
  18. Spark Rack will not silently or retroactively convert previously collected information into AI-training data.