OpenAI has released Privacy Filter, an open-weight model aimed at detecting and redacting personally identifiable information in text. The tool is designed to be context-aware and run locally, targeting “privacy-by-design” for enterprises and developers. The goal: protect sensitive data during AI training and processing without relying on cloud infrastructure.
India’s government is reportedly considering a limited exemption for early-stage startups from some compliance requirements under the proposed Digital Personal Data Protection (DPDP) bill. The rationale: easing pressure from data-handling obligations that could stifle startups while they develop data models and solutions. The DPDP draft also details strong penalties for breaches and proposes removing compensation provisions from the IT Act.
Your news, in seconds
Get the Beige app — every story in 60 words, updated hourly. Free on iOS & Android.
Meta is replacing its Accounts Center with a new Meta Account aimed at simplifying how users manage their experience across multiple Meta apps and devices. The change is designed to consolidate settings and make cross-app control more seamless, reducing the need to revisit different menus as you switch platforms within Meta’s ecosystem.
India’s customer service call recordings are often collected without genuine consent: customers are recorded for quality checks but have no real choice. As firms increasingly reuse this audio for analytics and AI, privacy risks grow. The editorial urges a clear audit trail for any mined data and stronger, e-commerce-like protections to ensure consent is meaningful.
Lovable, the app-building platform, has apologized after some public projects exposed chat data. The company says it wasn’t a data breach, but a combination of unclear product design and a technical mistake that let chat history appear publicly when projects were set to public. It has since changed default settings to private and claims the issue is fixed.
Instagram is testing a new “Instants” app designed for more fleeting sharing. Users can send disappearing photos to friends that are viewable only once, while the shared content stays accessible for up to 24 hours. The trial hints at Instagram leaning further into time-limited privacy and quick-hit social moments.
Never miss a story
Set alerts for the topics and sources you care about. Download Beige for free.
India’s current IT Act is widely seen as outdated and insufficient for today’s data protection needs. As technology, data collection, and online threats have evolved, the existing framework struggles to keep pace with modern safeguards. Analysts argue the country needs an urgent overhaul to better protect personal information and align with global expectations.
Adobe is expanding its AI ecosystem with a clear bet on first-party data to solve enterprise customer experience fragmentation. With privacy rules tightening and third-party cookies fading, companies increasingly need their own customer data unified and governed. Adobe Experience Platform is positioned as the backbone to keep compliance, control, and governance intact while powering scalable personalization.
Apple has released an urgent fix for a iOS logging flaw (CVE-2026-28950) that could leave deleted notification data on iPhones and iPads. The vulnerability reportedly came under scrutiny after the FBI recovered Signal messages from an iPhone, raising concerns that system-level weaknesses can bypass app-level encryption assumptions. iOS 26.4.2 and iPadOS 26.4.2 address the retained data issue.
OpenAI has released Privacy Filter, an open-source model that detects and redacts personally identifiable information before data ever leaves an enterprise environment. Built from a gpt-oss variant, it runs locally on laptops or in browsers and supports large 128,000-token inputs. The tool targets fast, high-throughput privacy pipelines, but comes with a caution against treating it as a full safety guarantee.
Reading on mobile?
Open Beige in the app for a smoother experience — free on iOS and Android.
Singapore has brought an AI-enabled teddy back on sale after a recall tied to concerns from a sex-chat scare. In trial runs, researchers from PIRG tested the toy’s conversation abilities and say that when they introduced the topic of “kink,” the AI engaged in ways that raised alarm over how these products handle sexual content.
Swipe through stories, personalise your feed, and save articles for later — all on the app.