Ethically sourced residential proxies are peer addresses whose owners knowingly agreed to share their connection, were compensated for it, and can withdraw at any time, with records that let an independent party check all three. Nearly every vendor uses the phrase; very few define it. The useful version is a set of testable criteria, each backed by a document or record you can ask to see.
This article sets out those criteria, contrasts vague claims with verifiable commitments, and explains how to tell marketing from evidence during an evaluation.
Why "ethically sourced" needs a definition
There is no industry standard, certification body or regulator-defined meaning for the phrase. A network that buries bandwidth sharing in clause fourteen of an app's license agreement can call itself ethically sourced, and so can one that shows every peer a plain-language notice and pays them. The words are identical; the supply is not.
That gap stopped being academic in July 2026, when a top-tier residential provider was seized by federal authorities after its network was found to be built on compromised devices. Its marketing had made the same kind of sourcing claims as everyone else's. For buyers, the lesson was that the claim itself carries no information. What a residential proxy botnet means for the buyer covers the exposure that follows when the claim turns out to be false.
The practical response is to define the term yourself, in criteria that can be checked, and apply them to every vendor on the shortlist, including your incumbent.
The eight criteria that make sourcing ethical
Each criterion below describes what should be true and the evidence that shows it.
1. Informed opt-in with a disclosure the user actually sees
The person using the device is shown, before anything happens, a plain-language explanation that their connection will carry other people's internet traffic, and they take an affirmative step to accept it. A clause in a license agreement does not meet this standard, and neither does a pre-ticked box or a notice shown after installation. Evidence: the exact disclosure text and screenshots of the consent flow, per partner application.
2. Compensation
The peer receives something of value in exchange, and the disclosure says what: cash, credit, premium features or the removal of advertising. Unpaid supply is a warning sign, because nobody has a reason to agree to it knowingly. Evidence: the compensation model per partner, stated in the disclosure the peer saw.
3. Easy reversal
The peer can switch sharing off from inside the application's own settings, without contacting support or uninstalling, and the address leaves rotation promptly once they do. Evidence: the location of the control, and a measured time from withdrawal to removal, ideally tested by an auditor.
4. No minors' devices
Partner applications that cannot enforce age gating, or that are aimed at children, are excluded. Evidence: a written exclusion in the supply-chain policy and the partner vetting criteria that enforce it.
5. Bandwidth, battery and data limits
Sharing should not degrade the peer's own use or cost them money. Typical conditions are idle-only operation, caps on bandwidth, and, on mobile devices, running only when charging and on Wi-Fi unless the peer has agreed otherwise. Evidence: the operating limits written into the SDK configuration and stated in the disclosure.
6. Partner vetting
Most peers arrive through partner applications, so the network is only as good as its partners. Each partner's disclosure and consent flow should be reviewed before onboarding and again periodically, and partners that change their flow without notice should be suspended. Evidence: the vetting procedure and the date of each partner's last review.
7. Audited records
An independent assessor samples addresses, traces each to its acquisition record, partner, disclosure and consent timestamp, tests withdrawal, and reports the exceptions found. Evidence: the attestation, its date and scope, and the exceptions list. What is IP provenance? describes what those per-address records should contain.
8. Prohibited channels in writing
A public, versioned supply-chain policy names both the channels the network permits and the channels it will not use: malware, bundled installers without knowing consent, compromised devices and routers, botnets, brokers who cannot evidence acquisition, and supply resold from other networks. Evidence: the policy itself, with a version and date.
Vague claim vs verifiable commitment
The difference between marketing and evidence is usually visible in the wording. A vague claim cannot be shown to be false. A verifiable commitment names something you can ask to see.
| Vague claim | Verifiable commitment |
|---|---|
| "Ethically sourced network" | A public, versioned supply-chain policy listing permitted and prohibited channels |
| "All peers opt in" | The disclosure text and consent-flow screenshots for each partner application, available on request |
| "Users are rewarded" | The compensation model per partner, stated in the disclosure the peer sees |
| "Peers can leave at any time" | A named in-app control, and a measured time from withdrawal to removal from rotation |
| "We work only with trusted partners" | A written vetting procedure, with each partner's last review date |
| "Compliant with applicable laws" | A DPA available before signature, and a named team that answers sourcing questions |
| "Independently verified" | An attestation naming the assessor, the date, the scope and the exceptions found |
| "We never use malware or botnets" | A trace of any address you choose back to its acquisition record |
The last row is the most important. Every other commitment can be documented and still be misleading; a trace of an address you selected at random from your own traffic is hard to fake.
How to tell marketing from evidence
Apply four tests to any vendor that sells ethically sourced residential proxies:
- Can it be falsified? If no possible observation could show the claim to be false, it is marketing.
- Is there a document? A claim backed by a named, versioned, dated document is a commitment. A claim that exists only on a web page is a slogan.
- Did a third party check it? Self-assessment is a starting point. An independent assessor who samples records, and reports the exceptions, is evidence.
- Will they trace an address you choose? A vendor with complete records can. A vendor that offers only pre-selected examples has chosen what you see.
Questions to send, copied as-is:
- Name every channel through which an address can enter your residential pool.
- Send the disclosure text and consent-flow screenshots for each partner application.
- How are peers compensated, and where is that stated to them?
- Where is the control that reverses consent, and how quickly does the address leave rotation?
- How do you exclude minors' devices?
- What limits apply to peer bandwidth, battery and mobile data?
- How are partners vetted, and when was each last reviewed?
- Who audited your sourcing, when, and may we read the exceptions they found?
- Send the current supply-chain policy, with its version and date.
- Trace these addresses from our trial traffic to their acquisition records.
Verifying a proxy provider's IP sourcing explains how to run the trace step and judge the answers, and the proxy provider due diligence checklist places sourcing alongside the security and contractual checks.
What ethical sourcing does not cover
Ethical sourcing is about how addresses enter the network. It does not describe what the network's customers do with them. A well-sourced network can still harm its peers if it carries abuse, so ask separately about customer screening and how the acceptable use policy is enforced.
Nor does it settle your own obligations. If you collect personal data through the network, the lawful basis, minimization and DPIA questions are yours, as GDPR web scraping and the proxy provider explains. Good sourcing makes the peer relationship in your assessment defensible; it does not make the collection lawful by itself.
Finally, it has a cost. Compensation, vetting, audits and exclusions all make ethically sourced residential proxies more expensive to supply. When a price sits well below the market, that is a reason to ask which of the eight criteria is not being met.
ProxyForge as a worked example
The ProxyForge sourcing page is organized around the evidence for each criterion rather than around claims. Residential peers join through a disclosed, compensated opt-in via partner apps and can reverse it. Our supply-chain policy is public and versioned, naming permitted channels and prohibited ones, including minors' devices. An independent assessor audits our sourcing twice a year by sampling addresses and tracing each to its acquisition record, and the attestation is available under NDA.
If you are comparing vendors of ethically sourced residential proxies, apply the table above to us as you would to anyone else. Send an address from a trial on our residential proxies, and we will return its full record.