社会ガイド
Data Brokers and AI Profiling
Data brokers assemble personal information from public records, purchases, apps, websites, and other sources, then may use statistical models to build or sell profiles.
このページでは3 分で読めます
概要
These profiles can infer interests or life events without a person directly providing them. The risks and consumer rights depend on the jurisdiction, data type, use, and applicable privacy or consumer-protection laws.
ディープダイブ
A data broker collects or derives information about people, often combining public records, commercial transactions, apps, websites, and other datasets. Brokers may sell records or inferred audience segments to marketers, campaigns, insurers, employers, lenders, or people-search services. A segment may describe an interest or predict an event; AI can scale these inferences, while conventional matching also remains common. The information used to build a profile may be incomplete, stale, or linked to the wrong person. A prediction such as “likely new parent” is an inference, not a fact about an individual. Profiles can still affect what advertisements, offers, prices, or opportunities a person receives. Whether a particular use is lawful depends on sector-specific and jurisdiction-specific rules; there is no single U.S. law that gives every person the same broker rights. California’s Delete Act is one example of a state intervention. The California Privacy Protection Agency launched DROP on January 1, 2026; consumers can submit one request to registered brokers, and brokers began processing requests August 1, 2026, accessing the system at least every 45 days. Deletion requests are subject to legal exceptions, and the platform covers registered brokers rather than every company holding data. Privacy laws can cover inferences as well as directly supplied details. California’s CCPA includes certain inferences in personal information, and its sensitive-information rules cover inferences used to create profiles reflecting specified sensitive categories. The EU GDPR also gives special protection to specified categories of personal data. The exact definition, legal basis, and consumer rights differ. A broker profile may fall under one rule for collection and a different rule when used for credit, employment, housing, or advertising. Consumers can inspect privacy notices, use opt-out or deletion tools where available, correct source records, and limit precise-location or ad tracking. Businesses should document sources, inference logic, downstream sharing, and rights processes.
戦略的影響
リスクと安全性
AI による壊滅的な被害も日常的な被害も、誰がリスクを理解し、誰が行動できるかにかかっています。
より明確な判決
国民と専門家のリテラシーは、強力な安全政策が政治的に可能かどうかを左右します。
誇大広告を打ち破る
明確な説明は、誇大広告、研究室の PR、曖昧な倫理劇場に囚われることを減らします。
The Future of Data Brokers and AI Profiling
California’s DROP process is operational for registered brokers, but statutory exceptions remain and other jurisdictions use different rights and definitions. A consumer’s available remedy therefore depends on where they live, which entity holds the profile, and how it is used. New data sources can also support sensitive predictions that a company never collected directly. Before directing someone to a request mechanism, check the current agency registry and governing law. For product teams, review notices and downstream sharing when a new inference feature changes the profile’s purpose. Registry and platform procedures may be updated, so cite the date checked.
現実世界の実装
A retailer matches loyalty records with broker data to estimate household interests and choose which customers receive a promotion.
A campaign uses modeled scores to prioritize outreach, while the data source and prediction remain invisible to voters.
A people-search profile combines addresses and relatives; an advocate helps a client use available removal and safety procedures.
A California resident submits one deletion request through the state’s DROP platform, which registered brokers began processing August 1, 2026 subject to statutory exceptions.
リスクとガードレール
能力が複雑になる一方で、実存的なリスクを SF として扱います。
高度な自律性の下での調整による表面製品の安全性を混乱させる。
英語以外や専門家ではない聴衆には、低品質の情報源しか提供されません。
実装ロードマップ
製品の危害、誤使用、制御不能/調整不良のリスクを分離します。
どのような証拠がタイムラインと重大度についてのあなたの見方を変えるかを尋ねてください。
マーケティング上の主張よりも、一次情報源と具体的な評価を優先します。
意識だけでなく、キャリア、政策、資金、スキルなど、行動経路を 1 つ特定します。
探検を続けましょう
Free newsletter
Get the daily AI briefing
Three verified AI stories every weekday morning, written in plain English. Free forever, no ads.
One email each weekday. Unsubscribe in one click. We never sell or share your address.
Test yourself
Take the Data Brokers and AI Profiling quiz
Instant feedback on every answer, and a shareable certificate with a verifiable ID once you pass a course.
Support free AI education. AI Understanding is a 501(c)(3) nonprofit — no ads, no paywall, ever. Make a donation
よくある質問
What is Data Brokers and AI Profiling?
Data brokers assemble personal information from public records, purchases, apps, websites, and other sources, then may use statistical models to build or sell profiles. These profiles can infer interests or life events without a person directly providing them. The risks and consumer rights depend on the jurisdiction, data type, use, and applicable privacy or consumer-protection laws.
What can a data broker sell besides directly collected records?
Brokers may sell profiles, audience segments, or predictions built from multiple sources.
Is a profile label such as “likely new parent” a verified fact?
The guide distinguishes a prediction from a known fact about an individual.
When did California residents gain access to the DROP request platform?
The California Privacy Protection Agency launched DROP for consumers on January 1, 2026.
When did registered brokers begin mandatory DROP processing?
The Agency says data brokers must begin accessing and processing requests August 1, 2026.
How often must data brokers access DROP to retrieve requests?
The Delete Act requires access at least once every 45 days beginning August 1, 2026.
学び続ける
関連ガイド
このトピックのために選ばれたその他のガイド