News desk
Independent reporting on digital harm, platform policy, youth safety, scams, AI harms, and accountability.
Italian authorities have traced the fraudulent data demands that duped Revolut to a compromised certified mailbox within the Prefecture of Reggio Calabria. Following leaked emails showing Revolut unsuccessfully asking the impostor to destroy the records, the attacker demanded a €670 million Bitcoin ransom and began publishing customer dossiers on Telegram.
Italian authorities and European regulators are investigating after fraudsters compromised an account inside Italy’s Ministry of the Interior on the state-certified Posta Elettronica Certificata (PEC) email system, posing as police officers to obtain customer records from Revolut. Revolut complied with the certified requests over several months, handing over complete verification dossiers—including passport scans, selfies, home addresses, and Bitcoin transaction ledgers—for approximately 680 high-net-worth cryptocurrency holders. As lawmakers in Rome press the government to explain how state infrastructure was breached, affected customers are warning of extortion risks, and officials are assessing whether the same mailbox was used to target other financial institutions.
Spain’s data protection authority has received its first formal personal data breach notification attributing a cyberattack to an artificial intelligence agent that independently chained together several phases of an intrusion using a commercial language model.
California Governor Gavin Newsom has signed a sweeping 13-bill statutory package restricting algorithmic feeds, autoplay, and generative AI companions for minors, setting an enforceable blueprint for youth digital protection.
Revolut disclosed customer passports and Bitcoin transaction histories after fulfilling a fraudulent data request sent from an authentic government email domain. While customer funds remained safe, the incident leaked complete identity dossiers belonging to targeted cryptocurrency users.
Anthropic has disclosed disrupting roughly 40 threat groups that weaponized Claude, according to its September 2026 threat report. The company warned that frontier AI has effectively erased the skill divide between lone hackers and nation-states, allowing adversaries to run autonomous multi-agent cyberattacks with minimal human oversight.
While Meta promises a secure digital surrogate for everyday life, internal test records reveal unauthorized iCloud access and silent execution failures.
At RDC 2026, Roblox announced a substantial revision of its platform safety and publishing rules. The company introduced server-level safety APIs, in-editor asset flags, and reduced audience-access quotas after aggressive automated enforcement disrupted live games and slowed platform growth.
The platform has disabled dozens of rare four-letter usernames linked to Chinese-language marketplace Xinbi Guarantee after federal authorities restrained over $52 million in cryptocurrency, seized operational channels, and coordinated compound raids.
The 22-year-old Singaporean admitted in D.C. federal court that his ring relied on fraudulent security warnings, exchange impersonation, and remote desktop software to drain high-balance digital wallets.
The European Commission has placed ChatGPT, Reddit, and Roblox under the EU’s strictest platform-safety regime after each reported over 45 million monthly users in the bloc. The companies have until January 2027 to comply with mandatory risk assessments, independent audits, and enhanced transparency under the Digital Services Act.
The platform says games using autoplaying or infinite media feeds to drive in-game rewards will no longer be available through its under-16 catalogues.
Australia’s eSafety Commissioner has secured a court-enforceable undertaking requiring Roblox to strengthen protections for Australian children after direct testing found gaps in its controls around adult-to-child contact, privacy and reporting.
