Privacy

Twitter

Facebook

Copy Link

Key Facts

Key LawsGDPR, CCPA, EU AI Act

Key TechniquesFederated Learning, Differential Privacy

Top News

Vercel breach traced to AI tool after employee grants unrestricted Google Workspace access

Vercel, the cloud platform behind Next.js, disclosed a security breach originating from Context.ai, a third-party AI tool. A Vercel employee granted the AI tool unrestricted OAuth permissions to their corporate Google Workspace account. When Context.ai was compromised through infostealer malware, attackers inherited those permissions and accessed Vercel's internal systems, exposing non-sensitive environment variables for a limited subset of customers.

Technology5 sources

Vercel breach traced to AI tool after employee grants unrestricted Google Workspace access

Technology5 sources

State AI Laws Multiply as Trump Administration Pushes National Standard Over Local Regulation

The Trump administration is waging a multi-front campaign to block state-level AI regulations, deploying legal threats and funding penalties to enforce a single national standard for artificial intelligence. But states aren't backing down—over 1,000 AI bills have been introduced across legislatures, with former tech employees turned lawmakers leading the charge for stronger local protections despite White House opposition.

Policy3 sources

State AI Laws Multiply as Trump Administration Pushes National Standard Over Local Regulation

Policy3 sources

Tinder adopts iris-scanning Orb to verify humans as AI bots proliferate across dating platforms

Sam Altman's World project announced a global partnership with Tinder, allowing users to verify their humanity through iris scans. The biometric identity company also expanded to Zoom and Docusign as platforms seek solutions to distinguish real people from AI-powered fake accounts. The move addresses growing concerns about romance scams and deepfakes, though privacy concerns persist.

Technology3 sources

Tinder adopts iris-scanning Orb to verify humans as AI bots proliferate across dating platforms

Technology3 sources

Failed startups are selling Slack chats and internal data to AI companies for up to $100,000

Shuttered companies are monetizing their digital remains—Slack messages, emails, and internal documents—to AI developers hungry for real-world training data. SimpleClosure has processed nearly 100 such deals in the past year, with payouts ranging from $10,000 to $100,000. But privacy advocates warn that workplace communications contain identifiable information that anonymization may not fully protect.

Technology3 sources

Failed startups are selling Slack chats and internal data to AI companies for up to $100,000

Technology3 sources

Perplexity Personal Computer AI for Mac launches with advanced agent control and security

Perplexity has launched Personal Computer, an AI assistant for Mac users that manages local files, native applications, and web browsing. Available to Perplexity Max subscribers at $200 per month, the system uses 20 AI models to execute complex tasks across apps like iMessage and Calendar, with security features including two-factor authentication and a kill switch.

Technology2 sources

Perplexity Personal Computer AI for Mac launches with advanced agent control and security

Technology2 sources

Samsung faces user backlash over Galaxy AI in heated Reddit AMA as fans call features a 'burden'

Samsung held a Reddit AMA where users questioned the value of Galaxy AI features, calling them burdensome and unnecessary. VP Annika Bizon addressed concerns about data privacy, feature utility, and customization options, revealing the company's vision for AI to become invisible and effortless like electricity.

Technology2 sources

Samsung faces user backlash over Galaxy AI in heated Reddit AMA as fans call features a 'burden'

Technology2 sources

Lawmakers voice existential fears about AI at congressional roundtable on American power

A House Oversight Committee roundtable on artificial intelligence turned existential as lawmakers from both parties aired deep anxieties about the technology. From concerns about AI-generated pornographic images to fears of engineering our own destruction, members of Congress pressed experts on cybersecurity threats, military applications, and whether the institution can implement common sense guardrails before it's too late.

Policy4 sources

Lawmakers voice existential fears about AI at congressional roundtable on American power

Policy4 sources

Tech enthusiasts build local LLM servers on Raspberry Pi and phones, proving on-device AI works

Developers are successfully running local LLM servers on modest hardware like Raspberry Pi 5 and smartphones, eliminating subscription fees and privacy concerns. Google's Gemma 4 models have made on-device inference practical, with users reporting speeds of 5-8 tokens per second on single-board computers and phones. The shift challenges assumptions about the hardware requirements for self-hosted AI.

Technology7 sources

Tech enthusiasts build local LLM servers on Raspberry Pi and phones, proving on-device AI works

Technology7 sources

Mozilla launches Thunderbolt AI client targeting enterprises seeking data sovereignty

Mozilla unveiled Thunderbolt, an open-source enterprise AI client that lets organizations run self-hosted AI infrastructure without relying on cloud providers like OpenAI or Microsoft. Built on the Haystack framework, it promises data sovereignty, end-to-end encryption, and freedom from vendor lock-in while supporting multiple AI models and platforms.

Technology5 sources

Mozilla launches Thunderbolt AI client targeting enterprises seeking data sovereignty

Technology5 sources

EU forces Google to share search data with rivals and AI chatbots under Digital Markets Act

The European Commission has issued preliminary findings requiring Google to share key search data with competitors under the Digital Markets Act. The measures explicitly include AI chatbots as data beneficiaries, sparking fierce pushback from Google over privacy concerns. Brussels aims to level the playing field in search markets, while Google claims the demands exceed the DMA's mandate and could jeopardize user privacy.

Policy2 sources

EU forces Google to share search data with rivals and AI chatbots under Digital Markets Act

Policy2 sources

ChatGPT could soon remember your face to streamline AI image generation for social media trends

OpenAI is testing a new feature that allows ChatGPT to remember user-uploaded images, eliminating the need to re-upload photos repeatedly when creating AI-generated images for viral social media trends. Spotted in the ChatGPT Android app beta, the 'Reference photo' option lives in the Memories section and gives users full data control to replace or delete stored images anytime.

Technology2 sources

ChatGPT could soon remember your face to streamline AI image generation for social media trends

Technology2 sources

Claude AI users must now verify identity with government ID as Anthropic introduces new checks

Anthropic has begun rolling out identity verification requirements for Claude AI users, requiring government-issued photo IDs and live selfies for certain capabilities. The move, handled through third-party partner Persona Identities, has sparked user privacy concerns despite assurances that verification data won't be used for model training. The decision marks a shift for a platform that attracted millions fleeing OpenAI over surveillance fears.

Technology4 sources

Claude AI users must now verify identity with government ID as Anthropic introduces new checks

Technology4 sources

Myseum stock surges 181% after AI rebrand, following Allbirds in speculative pivot wave

Myseum jumped 181% after announcing its rebrand to Myseum.AI, a day after Allbirds soared 582% on its AI pivot. The social media firm is developing localized, privacy-first AI agents for managing personal media. But the surge echoes past speculative manias, raising questions about whether struggling companies are capitalizing on investor enthusiasm at shareholders' expense.

Business3 sources

Myseum stock surges 181% after AI rebrand, following Allbirds in speculative pivot wave

Business3 sources

Security researcher exposes new Windows Recall flaw, but Microsoft says it's not a vulnerability

A cybersecurity researcher has developed TotalRecall Reloaded, a tool that can extract decrypted data from Windows Recall by exploiting the AIXHost.exe process. Despite responsible disclosure, Microsoft claims the access patterns are consistent with intended protections and do not constitute a security vulnerability, sparking debate about whether features that record user activity can ever be adequately secured.

Technology5 sources

Security researcher exposes new Windows Recall flaw, but Microsoft says it's not a vulnerability

Technology5 sources

AI Agents Hijacked via Prompt Injection: Bug Bounties Paid, Security Advisories Withheld

Security researchers exploited prompt injection vulnerabilities in AI agents from Anthropic, Google, and Microsoft, stealing API keys through GitHub Actions integrations. All three companies paid bug bounties ranging from $100 to $1,337 but issued no CVEs or public advisories, leaving users on older versions exposed to potential attacks.

Technology4 sources

AI Agents Hijacked via Prompt Injection: Bug Bounties Paid, Security Advisories Withheld

Technology4 sources

Prev 1…282930…69 Next