Why Public Proxies Are Dangerous: Risks, Speed, Tampering

A public proxy has no passport: it is an address from a table, behind which sits someone's server with unknown intentions. Sometimes a kind enthusiast stands behind such an address, more often someone's leaky configuration, and occasionally a trap set on purpose. The problem is that from outside these options look identical. Let's go through the five risks this creates and ways to check each one before it costs you money.

Risk 1: strangers watching your traffic

The whole stream passes through a stranger's machine. Encrypted sites protect the content, but the node owner still sees domains, times and volumes: enough for a profile of your interests. Any unencrypted site is fully open, with its forms and cookies. Hence the iron rule: never log in to accounts, enter codes or open payment pages through a public address. This rule knows no "just a quick look" exceptions.

Risk 2: tampered content

A middleman rewriting traffic can change it as well as read it: insert its ads or a data-collecting script, swap a wallet address on an unencrypted page or answer with a fake login page. Browser certificate warnings are the last line of defense here: if a site suddenly complains about its certificate through a proxy, close the tab at once, as this is no glitch. To check, open our anonymity check through the line and compare headers with what you expect, since traces of tampering often show up right there.

Risk 3: an address exposed everywhere

Users are not the only readers of public lists, as platforms read them too and feed the addresses into filters in batches. So the typical fate of a public address is a search captcha, a marketplace ban and silence from mail services. Run the address through RBL blacklists: dozens of entries are normal for a public address, and with such a reputation real work on it makes no sense.

Exposure also has a money side: a task run on such addresses wastes resources. A scraper gets captchas instead of pages, a checker reports false failures, and the results cannot tell target site problems from address problems. Any analytics collected through a public list needs rechecking, meaning repeat work.

Risk 4: zero speed and stability

Hundreds of people share an open address at once, its channel is a home-grade one, and it lives for hours. For a one-off page view that is tolerable, but for any regular task it generates timeouts and false errors. We cover this diagnosis in the article on free proxies that do not work, along with a symptom table.

Risk 5: other people's business on your address

Unknown people with unknown tasks work through the same address alongside you. Their spam and attacks become the reputation of the address you exit from: sites do not investigate which node user misbehaved. And if the node is someone's hacked server, you are technically using another's resource without the owner's knowledge, and the legal side of such savings is murky in most jurisdictions.

How addresses get into lists

Understanding the pipeline explains the product's quality. Scanners constantly comb address spaces and try common ports: whatever answers goes into a database. Aggregators reprint each other's databases, so one address wanders across dozens of sites for weeks after its death. The further a list is from the source, the higher its share of dead and dangerous lines. Hence a practical filter: only a list with its own checks and fresh timestamps has value, and the rest is archaeology.

A separate note in case you have already used public addresses and entered data: change the passwords of services you logged in to, and check your email for unknown sessions and forwarding filters. If payment details were among the things you entered, reissuing the card is easier than watching it for months. Paranoia here costs less than recovery, and we lay out the full procedure in the article on account hacks.

Where public addresses make sense

There is one zone: anonymous one-off tasks. Debugging code, checking a page's availability, learning experiments. For these we keep a list with ping and check date for each address. Anything involving accounts, money, regular data collection and results you will answer for belongs on addresses with a known operator: a contract and support stand behind the line there, and no random server. This line runs between "nothing to lose" and "something to lose" rather than between free and paid: once the second appears, public lists close. We gathered a format comparison on uptime, speed and support in a table in our review of paid and free proxies, which helps explain this line to colleagues and to the budget.