AI Terms & Conditions

1. AI-Generated Content — Disclaimer

Savage Lex AI sends sanitized requests through OpenRouter, which routes them to selected third-party language models. Model developers may include OpenAI, Google, DeepSeek, Anthropic, and others as our configured models change. AI-generated content is:

• NOT legally reviewed or approved by any attorney
• NOT guaranteed to be accurate, complete, or applicable to your jurisdiction
• NOT a substitute for consultation with a licensed attorney
• Generated by automated systems without human legal judgment
• Subject to the limitations, biases, and error rates of AI language models

You assume all risk for reliance on AI-generated content. Savage Ventures LLC does not warrant the accuracy, completeness, or legal sufficiency of any response.

2. Model Accuracy & Limitations

AI language models have known limitations. AI responses may:

• Contain factual errors, hallucinations, or outdated information
• Misapply law to your specific facts or jurisdiction
• Fail to recognize important nuances, defenses, or procedural requirements
• Be based on training data that does not include recent legal changes in your state
• Produce different outputs for identical inputs (non-deterministic)
• Use probability-based language prediction rather than legal reasoning

Always independently verify AI-generated information using official legal sources, court websites, or a licensed attorney. Do not rely solely on Savage Lex AI for legal decisions.

3. Third-Party AI Providers

Savage Lex AI transmits sanitized user input through the following services:

• OpenRouter — the routing service used for conversational AI and live-web verification
• Model developers selected through OpenRouter, which may include OpenAI, Google, DeepSeek, Anthropic, and others
• Mistral AI — document text extraction when that upload feature is enabled

You acknowledge and consent to transmission of sanitized input to these services. Current model selections may change for reliability, safety, or quality. Each provider has separate terms and privacy policies. Review their policies at:

• OpenRouter: https://openrouter.ai/terms
• Mistral AI: https://mistral.ai/terms

Savage Ventures LLC does not control, endorse, or assume liability for third-party AI providers' actions, policies, or breaches. Any breach, abuse, or misuse of your data by third-party providers is their responsibility, not ours.

4. Automated Redaction — NOT Guaranteed

Before transmitting input to third-party AI providers, Savage Lex AI applies automated redaction to attempt to remove personally identifiable information (PII). This process is:

• Automated, not human-reviewed
• Best-effort only — NOT guaranteed to catch all PII
• Based on pattern matching and regex rules, not semantic understanding
• Subject to false negatives (missing PII) and false positives (over-redaction)

CRITICAL: DO NOT RELY ON AUTOMATED REDACTION AS YOUR ONLY PROTECTION. You must manually avoid entering real names, case numbers, addresses, SSNs, or financial information.

Even if information appears to be redacted, it may still be transmitted to third-party AI providers. Assume any input you submit may be seen by third parties. Savage Ventures LLC is not liable for any PII transmitted despite redaction efforts.

5. Training Data Use

For Guest and Free Plus service, Savage Ventures LLC may use sanitized chat interactions for quality evaluation, redaction improvement, and training or improving Savage Lex systems. Raw uploads, raw OCR text, and unredacted private facts are not included in training datasets.

Lite and Standard chat requests are excluded from Savage Lex training datasets and are routed with a provider-side data-collection denial. Provider availability may be reduced when that privacy requirement is applied.

Original chat messages are retained separately in a restricted Raw Vault under the Privacy Policy. Raw Vault contents are not sent to model providers or included in training exports.

Guest and Free Plus users may opt out of future Savage Lex training use by emailing privacy@savagelex.com. A deletion request removes eligible records from future datasets, but cannot reverse a training run already completed.

Aggregated and anonymized data may be retained indefinitely for research and product improvement.

6. Conversational Context

Savage Lex AI may use sanitized recent conversation history so follow-up questions can be understood. Depending on the plan and feature, the system may:

• Replay a limited number of recent chat messages
• Store sanitized chat history for saved sessions
• Use user-provided, sanitized memory or case context in a later prompt
• Lose or omit older context when a plan or model context limit is reached

Conversation history does not make Savage Lex an attorney and does not mean the system understands every fact or document in your matter. Review each answer for missing context.

7. Legal-Information Freshness and Live Search

Savage Lex uses a maintained legal-information library and may use live web search for time-sensitive questions such as current statutes, deadlines, fees, forms, hours, addresses, or court procedures. Live search is selective, may be unavailable, and is not a guarantee that an answer is current. Responses may still miss:

• Recent changes to your state's statutes or local court rules
• Recent court decisions or case law
• Changes to filing fees, court procedures, or deadlines
• Updates to federal regulations or executive orders
• Changes that occurred after the model's training data cutoff date

Always verify information with your state's official court resources, your state bar association, or a licensed attorney before relying on AI-generated legal information.

8. No Duty to Update

Savage Ventures LLC has no obligation to update AI models or reprocess your data when legal information changes. Once a response is generated, it is not automatically refreshed or updated to reflect new legal developments.

You remain responsible for independently verifying that information is current and accurate.

9. Prohibited Uses of AI Output

You may NOT use Savage Lex AI output to:

• Impersonate a licensed attorney or legal professional
• File documents in court falsely representing them as attorney-drafted or attorney-reviewed
• Sell, republish, or license AI output to third parties without consent
• Train competing legal AI systems using our AI output
• Submit AI output as your own original work in an academic or professional context
• Use AI output in violation of applicable law or professional responsibility rules

Unauthorized use of AI output may result in account termination and legal liability.

10. Accuracy Disclaimers in Generated Content

Savage Lex AI automatically includes disclaimers in all generated content reminding users that the output is not legal advice and should be verified with a licensed attorney.

You are responsible for reading and understanding these disclaimers. Failure to read the disclaimer does not waive your acknowledgment of these AI Terms.

11. Liability for AI Errors

Savage Ventures LLC is not liable for any damages, losses, or adverse outcomes resulting from:

• Inaccuracy or incompleteness of AI-generated responses
• Misapplication of law to your specific facts or jurisdiction
• AI model errors, hallucinations, or false information
• Reliance on outdated legal information
• Failure to independently verify AI output
• Third-party AI provider errors or service disruptions
• Adverse legal outcomes, court dismissals, or sanctions resulting from use of AI output

This liability exclusion applies to all damages, whether direct, indirect, consequential, or punitive. See the Disclaimer and Terms of Service for the full Limitation of Liability clause.

12. Model Changes & Service Disruptions

Savage Ventures LLC may:

• Change, upgrade, or replace AI models without notice
• Modify how redaction is applied to user input
• Change third-party AI providers without notice
• Restrict or disable AI services due to abuse, legal compliance, or cost
• Temporarily suspend AI services for maintenance

We will attempt to provide notice of major service changes, but do not guarantee advance notification. Continued use after any change constitutes acceptance of the new model or provider.

13. Compliance With Laws

You agree to use Savage Lex AI only in compliance with all applicable laws, including your state's laws and U.S. federal law.

You may not use the platform to:

• Engage in fraud, impersonation, or deception
• Harass, threaten, or intimidate any person
• Violate attorney ethics rules or unauthorized practice of law statutes
• Circumvent legal restrictions on the practice of law
• Transmit unlawful or defamatory content

Violation of this section may result in account termination and referral to law enforcement.

14. Feedback & Improvement

Savage Lex AI may request feedback on the quality and accuracy of AI responses. Feedback is used to improve the system and train future models.

By submitting feedback, you grant Savage Ventures LLC a perpetual, worldwide license to use your feedback to train models and improve products. Feedback is subject to the same redaction and privacy protections as your input data.

15. Termination for Misuse

Savage Ventures LLC reserves the right to suspend or terminate your access to Savage Lex AI if you:

• Violate these AI Terms or any other applicable terms
• Misrepresent AI output as legal advice from a licensed attorney
• Abuse the system or attempt to circumvent PII redaction
• Submit unlawful, defamatory, or malicious content
• Attempt to extract, reverse-engineer, or train models on AI responses

Termination is effective immediately upon notice, and you may lose access to your account and prior data.