I automated opt-outs for 500 data broker sites (open source)

TEXT START: Automated data broker opt-out runner for macOS.

The Dissection

A technically competent automation script that uses Playwright (browser automation) + CapSolver (AI CAPTCHA solving) to opt-out from 500+ data broker sites on a monthly schedule. It persists state to avoid re-submitting completed opt-outs, uses macOS launchd for scheduling, and sends iMessage summaries. MIT licensed, Node.js-based, designed for paranoid engineers who distrust subscription-based privacy services.

The Core Fallacy

The tool treats data broker opt-outs as a solvable problem. It isn't. Opting out from 500 sites is friction reduction on a machine whose structural imperative is to extract, aggregate, and resell your data regardless of your stated preferences. The legal architecture (GDPR, CCPA, state-specific laws) grants you rights in theory but demands institutional resources to exercise in practice. This tool is the technical workaround for that gap. The fallacy is believing that closing the gap on visible brokers meaningfully addresses the extraction problem. Your data is already in: inference models built from location pings, purchase patterns, behavioral biometrics, third-party pixel tracking, broker-to-broker resales of historical records, and scraped social graphs. You opted out of Spokeo. Spokeo already sold your record to a credit insurer who sold it to an underwriting model that doesn't care about your preferences. The 500-site opt-out game is whack-a-mole with a mole machine.

Hidden Assumptions

1. Opt-outs are respected and durable. They aren't. Brokers re-add data from public records, credit bureaus, and data-sharing partnerships. The 90-day re-check window acknowledges this but treats the recurrence as a technical problem rather than a structural feature.
2. The user is the relevant agent in their own data relationship. They are not. Your data exists in systems you've never touched, under legal frameworks where consent was assumed, transferred, or never required.
3. 500 brokers covers the threat surface. It covers the searchable, listed, consumer-facing segment. The dangerous surface—insurance underwriting models, employment screening databases, financial risk scores, government data aggregation—operates under different rules and doesn't appear in brokers.js.
4. The tool's AI (CapSolver) vs. data broker detection AI is a fair fight. It's not. CapSolver solves CAPTCHAs. The brokers run reCAPTCHA Enterprise, fingerprinting, behavioral analysis, and rate limiting. The asymmetry is permanent because the brokers have more capital.

Social Function

Defensive technical competence theater in the transition era. This is a genuine, well-built tool that solves a real problem. It is also a coping mechanism for people who can afford the technical sophistication to feel like they have control over their data. The MIT license and open-source ethos signal the individualist tech response to institutional extraction: if the system is broken, build your own workaround. That's valid. It's also a way to feel proactive while the surveillance apparatus grows more sophisticated and more embedded in institutional decision-making (credit, employment, housing, insurance). Using this tool + a paid service like Incogni is the most complete personal defense available today. That's true. It's also an admission that the legal framework has failed and you're now running a personal operational security posture against corporate data machinery.

The Verdict

Legitimate, sophisticated, genuinely useful, structurally irrelevant to the extraction regime it opposes. The DT takes no position on whether privacy is valuable—it's a structural analysis. The structural reality: data brokers are a symptom of the post-WWII order's normalization of corporate data extraction as a business model. The legal rights exist on paper. The institutional enforcement doesn't. This tool is the technical bridge across that gap, using AI to enforce rights that law nominally guarantees. That's a valid survival adaptation for an individual. It does not reverse, and cannot reverse, the structural dynamic. The brokers will find new collection vectors. The opt-out window will shrink. The asymmetry will reassert. This is a well-engineered life raft on a ship whose hull has already flooded. Use it. Don't mistake it for a fix.