Top stories
Revolut confirms customer data breach via fraudulent government agency requests
Revolut disclosed that an unauthorized third party obtained sensitive customer data by submitting fraudulent requests from a legitimate government agency email domain. The exposed information includes identity details, contact information, and potentially verification selfies, account statements, and transaction histories for a limited number of customers. This incident highlights vulnerabilities in data access protocols, even when dealing with seemingly legitimate government communications.
OpenAI's Astra AI model faces safety concerns over opaque internal processing
OpenAI's upcoming Astra AI model is raising concerns among researchers due to its use of a recurrent depth or looped transformer technique, which makes its internal reasoning less transparent than other frontier AI models. This opacity could make it harder to monitor for undesirable behaviors, despite OpenAI's stated efforts to implement additional safety protocols.
OpenAI CEO Sam Altman States 2026 IPO Would Be 'Ill-Advised'
OpenAI CEO Sam Altman stated that the company will not go public in 2026, citing ongoing AI safety discussions and the need for the business to be ready. This clarifies previous reports that suggested a 2026 IPO was a possibility, with Altman indicating a later public offering when societal conditions and the company's readiness align.
Trezor Discloses ShipMonk Breach Exposed Data of 67,000 U.S. Customers
Hardware wallet manufacturer Trezor announced that 67,000 U.S. customers had their personal data exposed due to a breach at its shipping provider, ShipMonk. This incident is significant because it adds to previously disclosed exposures and highlights the risks associated with third-party data handling, even after assurances of data deletion.
PaperCut warns of active exploitation of zero-day vulnerability in NG and MF software
PaperCut issued an urgent security advisory regarding a zero-day vulnerability in all versions of its PaperCut NG and PaperCut MF print management software, which is actively being exploited in attacks. The company released emergency patches and advised organizations to restrict access to web interfaces of Internet-exposed servers, as this vulnerability poses a significant risk to affected systems.
Apple May Launch Camera-Equipped AirPods This Fall, Codenamed B790
Apple may release a version of camera-equipped AirPods, codenamed B790, as early as next month, according to references found in iOS 27 code and internal roadmaps. This potential release precedes another rumored variant, B798, which is anticipated in 2027. The cameras could expand health features, including vision assistance for blind and low-vision users when combined with Apple's AI-powered VoiceOver.
Craneware Reports Data Breach Affecting US Hospitals and Pharmacies
Craneware, a UK-based software provider for over 2,000 US hospitals, reported a data breach involving employee and customer information. The breach resulted in the theft of significant data, impacting hospitals' billing and patient management services. The incident has been contained, with investigations ongoing.
Russian National Charged in US for Malware Campaign Targeting 80,000 Freelancers
Searzhudin Tamirlanovich Aktulaev, a 40-year-old Russian national, has been extradited to the US and charged with orchestrating a phishing campaign that infected 80,000 freelancers with TVRAT and DarkVNC malware between 2016 and 2017. Aktulaev used 255 fake accounts on an unnamed freelance employment platform to distribute malicious Excel attachments, gaining remote control and stealing data from victims' systems. This case highlights international efforts to prosecute cybercriminals.
Two Berlin State Ministries Disconnected from Government Network Following Security Breach
Two Berlin state ministries, responsible for urban development and mobility, have been isolated from the city government's IT network due to a security breach. This incident has disrupted internal communications and some public services, highlighting vulnerabilities in government IT infrastructure.
Phil Schiller Steps Down from App Store and Apple Events Leadership
Longtime Apple executive Phil Schiller has stepped down from his leadership roles for the App Store and Apple events. The App Store division will now be overseen by Eddy Cue's services team, while Nola Weinstein will lead Apple events. Schiller will retain his title as Apple Fellow and work on unspecified initiatives.
NYC Public Schools Ban AI Use for Students Up to Eighth Grade, Restrict Digital Devices
New York City public schools will implement a one-year moratorium on AI use for students in 2-K through eighth grade, effective for the 2026-2027 school year, impacting approximately 600,000 students. This policy also bans companion chatbots across all grades and restricts teachers from using AI for grading, while introducing limits on digital device screen time for younger students. The decision reflects a stance against early AI integration in education, emphasizing human connection and critical thinking development.
ChatGPT, Claude, and Grok AI platforms experience simultaneous outages
Major AI platforms including OpenAI's ChatGPT, Anthropic's Claude, and X's Grok are currently experiencing widespread outages. The simultaneous disruption across these services suggests a potential underlying issue with Microsoft Azure, which many of these platforms utilize for cloud services.
Instagram to demote unlabeled AI-generated profiles, renames "AI creator" label
Instagram is implementing new rules for profiles featuring AI-generated personas, requiring them to use a new "AI-generated profile" label. Accounts that fail to self-label will have their reach limited, meaning their content will not be recommended to non-followers in Reels and Explore. This change addresses user feedback regarding misleading AI content and aims to increase transparency on the platform.
Meta Launches Muse Voice Transcribe for Real-Time Multilingual Speech Recognition
Meta's Superintelligence Labs released Muse Voice Transcribe, a real-time audio perception model for speech recognition. It offers multilingual, streaming transcription with speaker diarization and is available for Meta AI on Mac, Muse Code, and developers via the Meta Model API. The model supports over 70 languages, distinguishes more than 20 speakers, and achieved a 3.1% word error rate on the AA-WER Streaming benchmark for English, surpassing other comparable models.
Uber lays off 3,300 employees, 10% of global staff, to streamline operations
Uber is laying off approximately 3,300 employees, representing 10% of its global workforce, as part of a restructuring effort to reduce management layers and focus investments on ride-sharing, delivery, and robotaxi divisions. This move aims to simplify the company's structure, which CEO Dara Khosrowshahi stated had become overly complex due to growth.
Hackers deliver malicious Virtualizor update via BGP hijacking
Hackers used BGP hijacking to redirect traffic for Virtualizor's update infrastructure, delivering malicious updates to a small number of installations. Softaculous, the vendor, released a new Virtualizor version and advised users to check for a malicious service and reset credentials.
Sality P2P Botnet Dismantled After 23 Years of Operation
The Sality peer-to-peer (P2P) botnet, active since 2003, has been disrupted through an international law enforcement effort involving the U.S. Department of Justice, Europol, Eurojust, and private partners like CrowdStrike and the Shadowserver Foundation. The operation, which took place on August 31, 2026, included a P2P sinkhole and domain seizures, effectively neutralizing a long-standing threat that infected over 15,000 devices and distributed various malware, including the EggJagger clipjacking tool.
Crypto Traders Value Chinese Chipmaker CXMT at $425B on Derivatives Platform Pre-IPO
Crypto traders on a decentralized exchange are valuing China's largest memory chipmaker, ChangXin Memory Technologies (CXMT), at $425 billion, significantly higher than its official IPO valuation. This speculative valuation, driven by offshore investors lacking direct access to the upcoming Shanghai listing, highlights the demand for exposure to major Chinese tech companies through alternative financial platforms.
SpaceX finalizes acquisition of AI coding startup Cursor
SpaceX has officially completed its acquisition of AI coding startup Cursor, following an initial deal announced in April. This acquisition provides Cursor with access to SpaceX's computing infrastructure, including a large fleet of GPUs, to further develop AI capabilities.
Boston Scientific reports cyberattack disrupting global operations and order processing
Medical technology company Boston Scientific experienced a cyberattack on August 25 that disrupted its IT systems, causing a network outage and affecting its ability to process and ship customer orders globally. The incident impacts a major medical device manufacturer, potentially affecting the supply chain for critical medical equipment worldwide.
Z.ai Releases GLM-5.3 with Enhanced Coding and Cybersecurity Capabilities
Chinese AI startup Z.ai launched GLM-5.3, an updated language model with significant improvements in long-horizon coding and cybersecurity capabilities, which reportedly identified a serious vulnerability in Cursor. The model's advancements come from scaling post-training rather than a new base model, highlighting the potential for existing models to gain new capabilities through further training.
ATF confirms "major incident" after Qilin ransomware group claims breach
The U.S. Bureau of Alcohol, Tobacco, Firearms and Explosives (ATF) confirmed a "major incident" involving a breach of one of its standalone systems, following claims by the Qilin ransomware group. The agency stated that its main enterprise network, eForms system, and other ATF systems were not affected, and operations remain uninterrupted.
Google Play to mandate reduced memory usage and seamless device migration by 2027
Google Play will enforce new requirements for Android apps by 2027, including reduced memory usage and automatic sign-in during device transfers. These changes aim to improve app performance and user experience on Android devices, particularly in response to hardware supply constraints affecting memory availability.
Stripe Reportedly Acquires AI Gateway Startup OpenRouter for Over $7 Billion
Stripe has reportedly finalized a deal to acquire OpenRouter for more than $7 billion, according to Bloomberg. OpenRouter provides a single access point for customers to select and use various AI models, preventing vendor lock-in.
Situational Awareness Hedge Fund Sells Public Portfolio After Significant Losses
Situational Awareness, an AI-focused hedge fund founded by former OpenAI researcher Leopold Aschenbrenner, sold most of its public stock portfolio to Ken Griffin's Citadel following substantial losses in AI stocks. The fund's assets under management reportedly dropped from $45 billion to $10 billion in July, highlighting the volatility of highly leveraged investments in the AI infrastructure sector.
Amazon increases prices for Echo, Fire TV, and Kindle products by up to 60%
Amazon has raised prices on its Echo, Fire TV, Kindle, and Eero devices by up to 60%, citing increased memory and storage component costs. This change affects several popular Amazon hardware products, making them more expensive for consumers.
Apple Proposes 5-15% Commission for Off-App Store Purchases Amidst Legal Battles
Apple has proposed a commission structure of 5% to 15% for purchases made through external payment systems in the US, following a court order in its ongoing legal dispute with Epic Games. This proposal came after the Supreme Court denied Apple's request to pause lower-court proceedings, which are determining the fees Apple can charge for off-App Store transactions.
Apple reportedly lays off 60 Vision employees, signaling shift in VR development
Apple has reportedly laid off at least 60 employees from its VR development team, the Apple Vision Group, according to AppleInsider. This move suggests a potential reprioritization within Apple's spatial computing initiatives, despite previous assurances of continued investment in visionOS and the Vision Pro.
Grok AI Vulnerable to Data Exfiltration via Encrypted Malicious Instructions
Researchers discovered a new prompt injection attack against Grok that uses encrypted malicious instructions to bypass guardrails and exfiltrate user data. This method exploits the LLM's inability to distinguish between trusted user input and harmful content, allowing it to steal chat data and personal information.
Ransom Cartel Creator Sentenced to 16 Years for Ransomware-as-a-Service Operation
Maksim Silnikau, the 40-year-old Belarusian creator and administrator of the Ransom Cartel ransomware-as-a-service operation, was sentenced to 16 years in prison in Virginia. Silnikau developed the ransomware and recruited affiliates to attack at least 18 companies globally between 2021 and 2023, providing them with tools and infrastructure for intrusions and ransom negotiations.
SafePal data breach exposes order information for 39,798 customers
Cryptocurrency hardware wallet provider SafePal reported a data breach affecting approximately 39,798 customers, exposing names, email addresses, shipping addresses, phone numbers, and purchase information. A threat actor is now claiming to sell this stolen data on a cybercrime forum. This breach could lead to targeted phishing and social engineering attacks against affected customers.
New 'ShieldBreak' Zero-Day Exploit Bypasses Microsoft Defender Patch, Grants SYSTEM Privileges
Security researcher Nightmare Eclipse released "ShieldBreak," a new zero-day exploit for Microsoft Defender that bypasses a previous patch for the RoguePlanet vulnerability (CVE-2026-50656). This exploit allows SYSTEM privileges on fully updated Windows 10, Windows 11, and Windows Server systems, highlighting an incomplete patch for a privilege escalation flaw.
DeepSeek Updates V4-Flash Model, Raises API Prices, and Launches V4-Pro
DeepSeek has released DeepSeek-V4-Flash-0731, an updated version of its V4-Flash model, now in public beta, which maintains its architecture but shows performance gains through post-training. Concurrently, DeepSeek launched its V4-Pro model with agent upgrades and introduced new API pricing with peak and off-peak rates, effective August 16, 2026, which will increase costs for both V4-Flash and V4-Pro models.
US Imposes Tariffs Up to 100% on Foreign Drones and Components
The United States has implemented new tariffs on foreign-made drones and their components, with rates up to 100% for certain categories. This action aims to strengthen national security, reduce reliance on overseas suppliers, and encourage domestic manufacturing of drones and parts.
Microsoft to Launch Unified Copilot "Super App" This Year, Consolidating AI Features
Microsoft CEO Satya Nadella confirmed the company will release an AI "super app" this year, integrating Copilot's chat, coding, and agentic capabilities for both consumer and commercial use. This move consolidates Microsoft's AI offerings into a single application, aiming to simplify access and user experience by merging existing Copilot and Microsoft 365 Copilot apps.
EU Mandates AI Content Labeling and Transparency Under AI Act
The European Union's AI Act now requires AI-generated images, audio, and text designed to appear authentic to be labeled as AI-made, effective August 2 for new systems and with an additional four months for existing ones. This regulation aims to combat deceptive content and misinformation, ensuring users are aware when interacting with AI or AI-created media.
FBI Warns of Cybercriminals Hacking Accounts to Steal Explicit Images for Extortion
The FBI has issued a public warning about cybercriminals hacking into social media and online accounts of adults and children to steal explicit images and videos. These stolen materials are then used for blackmail, sold on criminal marketplaces, or shared with other criminals to facilitate further sextortion, with student-athletes and young boys frequently targeted.
Suno to Implement Audio Watermarking and Download Limits for AI-Generated Music
AI music generation service Suno announced new tools including audio watermarking and fingerprinting technology, revised download policies, and updated community guidelines. These changes aim to identify AI-generated content, prevent misuse on streaming platforms, and address ongoing legal challenges from music labels and artist organizations.
Spotify to Label AI-Generated Artists and Exclude Them from Recommendations
Spotify will introduce "AI Persona" labels for artists whose identities are AI-generated, starting in mid-September. These labeled artists will be excluded from Spotify's editorial and algorithmic recommendations by default, aiming to differentiate human artists from AI creations and address the influx of AI-generated content on the platform.
Mozilla Replaces GPG Signing Key for Firefox and Thunderbird After Accidental Exposure
Mozilla has replaced the GPG signing subkey used for Firefox and Thunderbird Linux tarballs, RPM packages, and checksum files. The previous key was inadvertently committed in an unencrypted copy to a private GitHub repository. This change primarily affects users who manually verify GPG signatures or use Firefox RPM packages on older Linux distributions, who will need to import the new key and the old key's revocation.