DeepCat Community Disclaimer & User Agreement
Version 1.0 — Effective March 2026
1. PLATFORM ROLE & NEUTRALITY DeepCat
(the "Platform") is a specialized software interface designed to
facilitate library cataloging tasks. DeepCat operates
strictly as a neutral pass-through agent. It does not provide Artificial
Intelligence (AI) services directly; rather, it facilitates the transmission of
user-provided data to third-party Large Language Model (LLM) providers (e.g.,
Google Gemini, Meta Llama).
2. USER-PROVIDED CREDENTIALS (BYOK) Access to AI
functionality requires a user-provided API key. By using the Platform, you
acknowledge and agree:
- Financial
Responsibility: You are solely responsible for all costs, quotas, and
billing generated by your API key.
- Security:
You are responsible for the confidentiality of your credentials. DeepCat does not see, store, or manage your API keys
on its server.
- Unauthorized
Use: The Platform Owner is not liable for any unauthorized charges,
quota depletion, or security breaches occurring on your third-party
provider accounts.
3. DATA PRIVACY & ZERO-RETENTION POLICY To ensure
maximum privacy, the Platform utilizes a "Zero-Retention"
architecture:
- Local
Storage: API Keys, Library Codes, and User Preferences are stored exclusively
within your browser's local storage (device-specific).
- In-Memory
Processing: Images and Metadata are processed in-memory. No
user-uploaded images or AI-generated outputs are
persisted to the Platform’s server disks or databases.
- Session
Purge: Once the browser session is closed or the "Clear"
function is used, all session-specific data is purged.
4. PERFORMANCE, TOKENS, AND COST ACCEPTANCE
- Variable
Performance: You acknowledge that the speed, accuracy, and quality of
output are contingent upon the selected LLM and its current operational
state. The Platform Owner makes no guarantee regarding specific analytical
results.
- Token
Consumption Logic: You acknowledge that "token" usage (and
subsequent cost/quota deduction) is dynamically determined by (i) the resolution and complexity of uploaded images,
and (ii) specific functional requirements (e.g., MARC21 vs. Dublin Core).
- Financial
Risk: By initiating any "Process" or "Run"
command, you accept the cumulative token costs. The Platform Owner is not
responsible for unexpected financial charges resulting from high-volume
image processing.
5. LIMITATION OF LIABILITY &
"HALLUCINATIONS"
- Verification
Duty: AI models are probabilistic and may "hallucinate"
(generate incorrect facts). You have an affirmative Duty of Care to
verify all AI-generated metadata before importing it into any production
Integrated Library System (ILS).
- No
Liability: UNDER NO CIRCUMSTANCES SHALL THE PLATFORM OWNER BE LIABLE
FOR ANY DIRECT, INDIRECT, INCIDENTAL, OR CONSEQUENTIAL DAMAGES, INCLUDING
BUT NOT LIMITED TO CATALOG CORRUPTION, DATA LOSS, OR FINANCIAL CHARGES
RESULTING FROM AI OUTPUT OR PLATFORM DOWNTIME.
6. SERVICE SUSTAINABILITY & FEES
- Current
Status: DeepCat is currently provided as a
free community tool.
- Right
to Modify: The Platform Owner reserves the right to introduce a
fee-based subscription model in the future to offset hosting and
maintenance costs.
- Notification:
Users will be provided 30 days' notice of any change to the fee structure.
Continued use after such changes constitutes binding acceptance of the new
terms.
7. AI ETHICS, BIAS, & ENVIRONMENT
- Algorithmic
Bias: You acknowledge that LLMs are trained on massive datasets that
may contain historical, cultural, or linguistic biases. The Platform does
not curate or filter these outputs.
- Environmental
Impact: AI inference is computationally expensive. Users are
encouraged to utilize the Platform intentionally to minimize the energy
consumption associated with AI inference.
8. DATA TRAINING POLICY DISCLOSURE
- No
Internal Training: DeepCat does
not utilize your images, habits, or processed data to train any
proprietary internal models. Your professional input remains your own.
- Upstream
Training: While DeepCat does not train on
your data, you acknowledge that Upstream LLM Providers (e.g.,
Google, Groq, Meta) may utilize data sent via
API for model improvement according to their specific Privacy Policies.
Your relationship regarding upstream training is governed solely by the
provider's terms.
9. NO WARRANTY (AS-IS) THE PLATFORM IS PROVIDED "AS
IS" AND "AS AVAILABLE." THE PLATFORM OWNER DISCLAIMS
ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE IMPLIED
WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE
PLATFORM OWNER DOES NOT WARRANT THAT THE SERVICE WILL BE UNINTERRUPTED, SECURE,
OR ERROR-FREE.
10. INDEMNIFICATION You agree to indemnify and hold
harmless the Platform Owner from any claims, damages, or expenses (including
legal fees) arising from your use of the Platform, your violation of these
terms, or your infringement of any third-party intellectual property rights.
11. GOVERNING LAW This Agreement shall be governed by
and construed in accordance with the laws of the jurisdiction in which the
Platform Owner resides, without regard to conflict of law principles.