Linzumi AI Terms
Last updated: June 7, 2026
These AI Terms apply to AI-assisted features in Linzumi's Service, including planning, drafting, code generation, review, summarization, image generation, document generation, agent workflows, and related features.
1. AI Inputs And Outputs
"Inputs" are prompts, files, repository data, project content, chat messages, instructions, metadata, images, audio, or other materials provided to AI features.
"Outputs" are text, code, images, plans, comments, reviews, summaries, diffs, suggestions, or other materials generated by AI features.
Inputs and Outputs are Customer Content under the Terms of Service.
2. Ownership
As between Customer and Linzumi, Customer owns Inputs. Subject to the Terms and applicable law, Linzumi assigns to Customer any rights Linzumi has in Outputs generated for Customer through the Service.
AI Outputs may not be unique. Other users may receive similar or identical outputs from similar prompts or context.
3. Human Review Required
AI Outputs may be inaccurate, incomplete, insecure, biased, outdated, non-deterministic, or unsuitable for Customer's intended use.
Customer is responsible for human review, testing, validation, security review, license review, and approval before using, publishing, deploying, merging, or relying on Outputs.
4. Code And Repository Outputs
Before merging, deploying, or relying on AI-generated code or repository changes, Customer should:
- Review the diff.
- Run tests and security checks.
- Confirm licensing and dependency impact.
- Confirm no secrets or sensitive data were introduced.
- Confirm the change complies with Customer policies and repository rules.
Linzumi is not a substitute for code review, security review, legal review, or production release controls.
5. No Professional Advice
AI Outputs are not legal, financial, medical, security, tax, or other professional advice. Customer should consult qualified professionals where appropriate.
6. Provider Processing
Linzumi may use third-party AI providers and external AI/token services to process Inputs, generate Outputs, meter usage, issue or manage tokens, and operate AI features. Linzumi may reach these providers directly and/or through model-routing providers such as OpenRouter, and reserves the right to add, remove, change, or route between providers as the Service evolves. By using AI features, Customer authorizes Linzumi to route Inputs, Outputs, Customer Content, and related metadata to the providers and external services Linzumi selects for those features. Providers may include OpenAI, OpenRouter, Anthropic, Google/Gemini, Cerebras, Groq, Wafer, and future providers listed on the subprocessor page at https://serve.linzumi.com/subprocessors.
Unless otherwise agreed in writing, Linzumi does not use Customer Content to train Linzumi foundation models.
Inputs may include source code, repository content, project files, chat messages, instructions, audio submitted for transcription, logs, issue or pull request metadata, and other Customer Content when needed to provide requested AI features. Each provider's own terms control Linzumi's use of that provider.
7. Prohibited AI Uses
Customer may not use AI features for prohibited conduct described in the Acceptable Use Policy, including malware, credential theft, unauthorized system access, deceptive impersonation, fraud, harassment, or unlawful high-impact decisions.
8. Safety Controls
Linzumi may apply filters, rate limits, context limits, model routing, logging, refusal behavior, abuse detection, and human review workflows to operate and protect AI features.
9. Experimental Features
AI features may be beta, preview, or experimental. Linzumi may change models, prompts, workflows, evaluation methods, or feature availability at any time unless otherwise agreed in writing.