Ethical Stack.

Kimi

Consumer AI assistant for chat, document work, and agentic tasks.

Score calculation

(2 + 4 + 2 + 1 + 7) ÷ 5 = 3.2

5 of six dimensions rated. Unrated dimensions are excluded from the average.

Full scoring method ↗
Setup details

Kimi consumer web and mobile application; excludes Moonshot API and Kimi Code CLI.

What fits your needs

Kimi consumer web and mobile application; excludes Moonshot API and Kimi Code CLI.

Use and access

This entry covers Kimi consumer web and mobile application; excludes Moonshot API and Kimi Code CLI.. Check whether your use is covered by the listed consumer or developer interface.

See the evidence
Tradeoffs to expect

Data route

A hosted request is processed by the service and may pass through an upstream model host; confirm the route and terms before sending sensitive material.

See the evidence
How to switch

Test first

Try a routine, non-sensitive task and verify current account, model, and route settings before moving ongoing work.

See the evidence
Practical questions

Terms and controls

Review current privacy, service terms, pricing, endpoint, and model version for the account you plan to use.

See the evidence

Energy practices

2/10↑

For the Kimi consumer web and mobile app boundary, public documentation establishes the service or route but not its request-level physical footprint. The consumer sources reviewed do not document a user-facing token cap or model/effort selector that could be scored as a workload control. Electricity sourcing is not mapped to this route.

See the evidence

This practices/transparency assessment covers the Kimi consumer web and mobile app boundary. The sources establish how the service or endpoint is exposed and don’t confirm a documented resource-saving control; they do not give a measured per-request energy, emissions or water value. The disclosure point reflects the stated operating boundary, not a product footprint, and company-wide electricity reporting is not attributed to this route.

1 + 1 reporting + 0 electricity + 0 controls = 2/10. This grades documented practices and transparency.

The reviewed sources don’t confirm request-level electricity, emissions or water; the actual serving facility and its power mix are also not mapped to Kimi consumer web and mobile app.

Reporting1/3
The reviewed documentation defines the Kimi consumer web and mobile app boundary. It gives an operating-scope description, not a dated quantitative energy, carbon or water report for this route.
Electricity sourcing0/3
No reviewed source maps Kimi consumer web and mobile app to a disclosed electricity mix or time-and-location-matched clean-power supply. The route's actual electricity use and emissions therefore remain unestablished.
Resource controls0/3
The consumer sources reviewed do not document a user-facing token cap or model/effort selector that could be scored as a workload control.

Water disclosure

Water reporting has not been assessed for this choice.

Training data

—Depends on the model↑

Kimi app model and version can change; the app terms and OpenPlatform policy do not expose a complete model-specific dataset and rights manifest for each request.

See the evidence

The consumer policy says de-identified inputs and outputs may be used to optimize models and provides a separate opt-out. That describes submitted chats, not the licenses, source inventory, or permissions for the model’s original training data; the model release must be assessed separately.

Exact model selection, release version, source inventory, permissions, post-training data, and creator-remedy scope remain unresolved.

Your data & privacy

4/10↑

Kimi’s consumer privacy policy covers the web and apps, says conversations and usage records are collected to provide and optimize services, and states that personal information is stored in China. Users can ask support to stop using content for model training.

See the evidence

The Kimi consumer policy applies to the website and apps, describes collection of conversation text, audio, images, video and files, permits use of de-identified content to optimize models, and states storage in China. It also gives a support-based training opt-out and chat deletion controls; those are meaningful controls but not a no-retention guarantee.

Chat content may be used for service/model improvement absent a support request, and data is stored in China; user controls exist but retention and opt-out timing are not specified.

The policy sets retention as necessary for service purposes rather than a fixed chat window; check feature-specific policies, account settings, and the current support process for opt-out requests.

Openness

2/10↑

This is a hosted service with a public interface but no documented self-hostable product runtime or weights in the reviewed sources.

See the evidence

The service provides a hosted product or API interface, while the reviewed product material does not expose the serving implementation and current model weights for independent deployment. This rates this service scope, not separately released model weights.

The score reflects service-level inspectability and control, not usefulness or model quality.

SDKs, APIs, or model families may have separate licenses and do not make this service itself self-hostable.

Worker protections

1/10↑

Moonshot’s reviewed Kimi terms, company site and model pages do not disclose workforce conditions, labor grievance mechanisms or supplier audit outcomes.

See the evidence

The sources reviewed provide product, terms, or corporate information but no qualifying disclosure about conditions for data annotators, evaluators, moderators, or infrastructure workers. This is a disclosure gap within the checked materials, not evidence that workers are mistreated or that no labor policy exists. These findings concern the developer or operator. This product’s use in a particular deployment needs separate evidence.

1 + 0 pay & conditions + 0 worker voice & remedy + 0 evidence & follow-through = 1/10. This grades documented worker safeguards, not a measured rate of mistreatment.

Worker counts, locations, pay, labor protections, worker voice mechanisms, supplier coverage, and independent audit or remediation outcomes were not established in the sources reviewed.

Pay & conditions0/3
Reviewed sources provide no qualifying disclosure of worker conditions; this zero is limited to the checked material and is not an exhaustive claim.
Worker voice & remedy0/3
No qualifying worker-voice mechanism or documented worker grievance outcomes were disclosed in the reviewed material.
Evidence & follow-through0/3
No independent labor audit results, worker-level corroboration, or remediation outcomes were disclosed in the reviewed material.
Workers rubric ↗

Military ties

7/10Not established↑

No Moonshot defense contract or military deployment is identified in the reviewed company and model sources.

See the evidence

Moonshot’s reviewed Kimi terms, company disclosures and model repository do not name a defense customer, contract or deployment. Third-party availability of its models doesn’t show Moonshot’s own military relationship; the public material reviewed does not supply a customer-complete accounting. These findings concern the developer or operator. This product’s use in a particular deployment needs separate evidence.

7 base + 0 safeguard adjustment = 7/10 (0 of 9 safeguard points; every 3 adds 1). No military relationship was established in the cited sources. This ranks above known defense ties without claiming a verified clean record.

Complete contract texts, undisclosed customers, end-use restrictions, actual workload deployment, and independent compliance or outcome reports are unavailable in the reviewed public material.

Use restrictions0/4
The reviewed terms or policy material contains no qualifying military-use restriction beyond general lawfulness/export compliance; this zero is limited to the reviewed source set.
Enforcement & oversight0/3
No military-specific review, enforcement, or remedy mechanism was disclosed in the sources reviewed.
Relationship disclosure0/2
No military-specific relationship or military-specific transparency policy was disclosed in the materials reviewed; this is a scoped disclosure finding.

The reviewed Kimi platform terms and Moonshot model materials do not identify a defense contract or military deployment. Model availability through third parties or permissive downstream deployment doesn’t confirm Moonshot’s own government or military customer relationship.

Military rubric ↗

BDS targets

Not listed↑

We did not find Moonshot AI among the named targets in the two official BNC campaign lists. This is a limited list check, not BDS approval.

See the evidence

Moonshot AI is not a named target in the checked lists, but BDS uses selective campaigns, so absence cannot establish a clean record. Running downloaded weights locally is a different decision from paying the developer or a cloud host; check the host as well. These findings concern the developer or operator. This product’s use in a particular deployment needs separate evidence.

This review covers the named targets in the BNC consumer/pressure guide and No Tech campaign, not every BDS campaign, local group, investment or supplier. Lists and ownership can change. A company’s absence from these lists is not an endorsement or proof that it has no relevant relationships.

More from Moonshot AI

All sources

Side by side