All articles tagged with #data leak
About 12 terabytes of Steam-era data (2003–2013) leaked, revealing game assets and prototypes from Valve titles such as an early Left 4 Dead build and data on the canceled Half-Life 2: Episode 3; the leak predates SteamPipe (2013) and contains only game data, not player data, with no post-2013 content, and Valve has not commented.
Over a weekend, a public 'teraleak' dumped thousands of pre-2013 Steam2 builds, betas, and Valve assets—including Portal, Half-Life 2 and Left 4 Dead—plus third-party titles. The archive, organized into depots and cataloged publicly (no passwords), has sparked a global data-mining frenzy and raised questions about publisher relationships and whether all content is runnable; no hacking was involved, just an exposed data store.
A leaked 12TB archive of Valve development assets from 2003–2013 has surfaced online, reportedly including early Portal 2 builds, a Left 4 Dead beta, and material tied to the cancelled Half-Life 2: Episode Three. Researchers are parsing the files for clues about Episode 3 and other unreleased work; Valve has not commented. The leak is separate from Valve’s earlier CEVA Logistics breach.
Hundreds of Claude AI conversations were publicly accessible via search engines when users shared links, exposing personal and work information; the logs appeared across pages before Anthropic blocked indexing and took steps to restrict access. The company says sharing is voluntary and that public logs can be archived, while Google notes it doesn’t control what pages are public and indexing is set by site owners.
A cybersecurity researcher uncovered a Tribeca Festival data leak that exposed 666,369 records from 2019–2026, including public-facing business contacts and a 13,535-entry “contacts” folder with names, emails and phone numbers of stars like De Niro, Lawrence, Jolie, Scorsese and Coppola. The data was stored in a plaintext backup left in the festival’s database and was removed quickly after disclosure; experts warn such exposed data could enable AI-powered phishing and other attacks, highlighting the need for stronger backups and access controls.
A hacker-leaked set of MSG security documents shows the arena used facial recognition to flag LGBTQIA celebrities and track a transgender woman, assigning risk scores; the disclosure, published by ShinyHunters and reported by Wired, has spurred lawsuits while MSG denies wrongdoing.
Reuters reports Apple is concerned after Tata Electronics, its Indian supplier, was targeted by a cyberattack that leaked confidential Apple documents related to the iPhone 18 Pro, including supplier lists, component designs, and drop-test images. Apple is investigating with Tata and preparing longer-term security measures, as the breach could affect its supplier relationships.
Meta has paused its Model Capability Initiative, an employee-tracking AI program, after sensitive data collected from workers—such as private conversations and performance data—was inadvertently exposed to the entire company. Meta says safeguards were in place and no data was definitively accessed, but the incident has renewed scrutiny of its AI security and follows other recent AI-related breaches at the company.
Meta has paused its Model Capability Initiative, an internal AI training program that used staff keystrokes and interactions, after a leak exposed private conversations, performance data, and transcriptions across the company. Meta says there’s no confirmed improper access but is investigating as the SEV 2 incident prompts employee backlash and follows other security incidents at the company.
Lawmakers from both parties pressed CISA for answers after KrebsOnSecurity reported a contractor publicly posted plaintext credentials and AWS GovCloud keys to a GitHub account, triggering ongoing credential rotation and breach containment. Experts warn that exposed keys could enable access to code, CI/CD pipelines, and sensitive systems. CISA says it is rotating leaked credentials and coordinating with vendors, while lawmakers demand answers about internal policies amid leadership turnover and broader concerns about the agency’s security culture.
Trump Mobile confirmed a data exposure via a third‑party platform that left customer names, emails, mailing addresses, phone numbers, and order IDs exposed; the vulnerability was patched after reports by Coffeezilla and TechCrunch, and rumors of lower pre-orders followed as the T1 launch remains delayed amid controversy.
A hacker data dump from Rockstar Games leaked online, exposing earnings from Grand Theft Auto and Red Dead Redemption; in the day after the leak, Take-Two Interactive saw its stock rise about 2.6% to around $206.66 per share, boosting the company’s market value by roughly $1 billion to about $38 billion, though shares remain below their 52-week high amid market volatility.
A Fortune report reveals Anthropic's Claude Mythos, the company's most powerful AI model, was exposed by a data leak. The leak has analysts weighing potential IPO implications for Anthropic and causing volatility in AI/cybersecurity equities as investors weigh the model's advanced capabilities against security risks.
Fortune reports that Anthropic left internal materials for its upcoming AI model Claude Mythos publicly accessible, exposing roughly 3,000 assets and heightening cybersecurity worries about the model's potential capabilities.
Researchers demonstrated a prompt-injection attack against Google Gemini by embedding a malicious payload in a Google Calendar invite description. When the recipient asks Gemini about their schedule, the assistant executes the embedded instructions, creates a new event, and copies private meeting details into the event description, leaking sensitive data to the attacker. Google added mitigations after the disclosure, underscoring the need for context-aware defenses as AI assistants access calendar data.