Privacy Digest 14/26
Supreme Court Supports Privacy Protections For Cellphone Location Data
The US Supreme Court has strengthened privacy protections for cellphone location data, ruling that law enforcement needs probable cause and a valid warrant before conducting geofence searches. These searches can sweep up location records from every device within a specific area and time window, potentially exposing data belonging to large numbers of people with no connection to a crime. The Court found that individuals have a reasonable expectation of privacy in their phone's location records and rejected the idea that this information is necessarily shared voluntarily with tech companies. The ruling limits broad digital searches while still allowing more narrowly targeted geofencing.
cnet.com
Nearly A Million Passports Left Exposed Online
Anyone with a simple URL could access sensitive identity documents from several European countries, with no password or hacking required. The files contained photos, names, and identifying numbers collected for membership and age verification and reportedly remained publicly accessible for months. The incident highlights the lasting risks of storing identity data without basic access controls. Unlike a compromised password, a passport or driver's license cannot simply be reset, potentially leaving affected individuals exposed to identity theft and document fraud for years.
cambridgeanalytica.org
Distractions No More: Hide Web Clutter Across Sites Easily
Not every online distraction is an ad. Short video feeds, social sharing widgets, sign-in prompts, and browser promotions can still clutter pages and pull attention away from what you came online to do. Ghostery's new Distractions Removal feature gives you ready-made controls to hide common types of web clutter across supported websites. Enable a category once, and Ghostery automatically hides matching elements wherever supported, with no custom filters or manual setup required. Combined with Distraction Eraser for site-specific customization, the feature gives you more control over how the web looks and feels, helping create a calmer browsing experience shaped around your preferences.
ghostery.com
Will Tech Giants Ever Let Us Opt Out Of AI Search Features?
AI is rapidly transforming search from a tool for finding information into one that delivers ready-made answers, often without giving users or publishers much choice. Growing frustration over mandatory AI features has pushed some people toward AI-optional alternatives, while regulators are beginning to intervene. In the UK, new requirements aim to give publishers greater control over how their content is used in AI search and improve transparency. The debate raises a broader privacy question: as people share increasingly personal information with AI tools, should opting out be a basic choice rather than a feature platforms decide to offer?
fastcompany.com
Collateral Damage: Claude Mythos And The Privacy Risks Of AI
Personal communications, health records, financial data, and digital identities all depend on systems remaining secure. AI models capable of rapidly discovering and exploiting software vulnerabilities could put that protection under growing pressure, increasing the risk of unauthorised access and mass data exposure. Anthropic restricted access to Claude Mythos and Fable over security concerns, before making Fable publicly available again on June 1, although it remains unclear what restrictions apply. Limiting access to powerful models may slow misuse, but it cannot solve the broader challenge of protecting personal data as AI capabilities continue to advance.
privacyinternational.org
Meta Halts Worker Tracking For AI Training Due To Privacy Fears
A company-wide program designed to record employees' mouse clicks, keystrokes, and computer activity for AI training has been paused after internal privacy concerns. Workers objected to the mandatory monitoring and questioned how the collected information would be stored and protected. The program was halted after some data was found to be potentially accessible to anyone inside the company, although there is no indication it was improperly accessed. Nearly 2,000 employees had reportedly signed a petition demanding the initiative be cancelled, highlighting growing tensions over workplace surveillance and consent as companies race to improve their AI models.
bbc.com