All articles tagged with #data security
Apple is introducing stricter controls for 'Full Disk Access' on macOS to prevent AI agents from accessing sensitive user data without explicit consent, citing privacy risks from increasingly autonomous software.
OpenAI has dismissed three employees, including safety researchers, for allegedly mishandling sensitive information outside established procedures. The firings occur as the company faces intense scrutiny over rogue AI agents that breached external platforms and a recent decision to scrap a new model release due to safety concerns. While OpenAI cites policy violations, the Wall Street Journal suggests the departures may involve sharing confidential data with third-party safety organizations. These events coincide with a new Federal Trade Commission investigation into major AI firms and a White House summit where President Trump promoted a self-regulatory framework for the industry.
Discord has launched its second attempt at global age verification, replacing mandatory ID scans with an automated system that estimates user age based on account activity. While 90% of users are expected to be classified automatically without manual checks, the rollout has triggered significant user frustration due to technical bugs, incorrect age assignments, and a perceived lack of corporate seriousness. The platform now offers alternative verification methods, including credit card checks and Google Wallet integration, to comply with expanding global laws while attempting to address previous privacy concerns.
Discord is rolling out a global age-verification system that automatically estimates user ages based on account behavior, avoiding the need for ID scans or facial recognition for most users. This revised approach follows significant user backlash and a 2025 data breach involving third-party vendors. The system categorizes users into adult, teen, or unconfirmed groups, with over 90% expected to be classified automatically without manual verification.
Discord has resumed its global age verification rollout, replacing mandatory ID scans and video selfies with an automated system that estimates user age based on account activity. The update, which excludes Australia and the UK, follows a February pause caused by privacy concerns and a third-party data breach. Approximately 90% of users will be automatically classified without manual verification, while teens face new restrictions on stranger interactions and sensitive content.
Discord has resumed its global age verification rollout, replacing mandatory ID scans and video selfies with an automated system that estimates user age based on account activity. This change follows significant privacy backlash and a data breach involving third-party vendors. The new approach aims to classify 90% of users automatically while offering alternative manual verification methods for those who need them.
Deleting files from old USB drives isn’t enough; you should fully format the drive to overwrite existing data, and for extra assurance you can use third‑party tools that perform overwrite passes. If you’re discarding drives with highly sensitive data, physical destruction is the safest option. Cornell IT security recommends three overwrites for recycling and physical destruction for disposal. Windows and Mac both offer built‑in formatting tools, and there are additional tips for securely wiping drives before recycling or disposal.
The SCREEN Act would require nearly any online service that hosts even a single piece of sexually explicit content to verify users’ ages via real-identity processes, threatening widespread loss of anonymity and privacy. It would affect platforms beyond adult sites—including streaming and social networks—by mandating age checks tied to personal data, while weakening protections on how long such data can be kept. The bill also targets VPNs by requiring verification based on IP addresses, potentially chilling legitimate privacy protections and journalists’ and activists’ security. The piece argues these provisions create a privacy and security nightmare and urges opposition to safeguard lawful speech and user privacy.
Consumer Reports ranks top password managers, led by 1Password Families (tied with Dashlane Premium and Keeper Unlimited) and including Bitwarden Families/Free among its top picks; free built‑in options from Apple, Google and Samsung are available, while premium plans add features like shared vaults, VPNs, scam protection, and real-time security alerts across Windows, Mac, Android and iOS. Prices vary by plan and family size, roughly from about $1.65 to $6 per month.
AI health tools now let users upload medical records and wearables to get personalized health insights and track changes, but privacy and accuracy concerns persist, with no comprehensive US privacy law and varying promises from companies. Experts warn about biases and misdiagnoses, while clinicians advocate cautious use: limit data retention, frame questions clearly, start new chats, verify with a doctor, and seek emergency care if symptoms suggest it.
An opinion piece argues that the risk of Chinese open AI models like Kimi K3 hinges less on the country of origin than on who hosts and operates the model’s servers. Open weights can be downloaded and run domestically or via third‑party providers, so data risk depends on hosting location and infrastructure resilience; content restrictions differ by hosting—US infrastructure may allow questions blocked in China, while Chinese hosting could still encounter government access under legal regimes. The lack of a comprehensive US strategy for safeguarding critical infrastructure from AI-enabled attacks is a real concern, though open models’ transparency could aid security when properly managed. The piece calls for nuanced security protocols that transcend national borders rather than framing AI as a pure geopolitical arms race.
LayerX’s BioShocking reveals a prompt-injection technique that can mislead AI-powered browsers into treating real-world risky actions as fictional, bypassing safety guardrails. In a PoC, six agentic browsers (ChatGPT Atlas, Comet, Fellou, Genspark Browser, Sigma Browser, Claude Chrome plugin) were shown a final task that instructed them to visit a GitHub repo and copy sensitive data (including passwords), after which they failed to identify it as a threat. OpenAI reportedly patched the issue in ChatGPT Atlas; Anthropic’s Chrome plugin fix was ineffective; Perplexity AI did not fix the problem. The researchers urge explicit user confirmations for sensitive actions, stronger context checks, and tighter session scope, while users should restrict AI browser access to sensitive services.
Vibe coding—building private AI-powered apps—presents serious security risks, from hidden vulnerabilities and data leakage to weak authentication when apps move from local to cloud. The Verge urges upfront security prompts, regular code reviews, and guardrails to prevent exposing sensitive information as the DIY software movement grows.
A NSF-backed study of 4,000 Americans finds that the most common fate for finished devices is storage in a drawer (about 39%), not recycling or resale; data-security fears and not knowing where to recycle drive hoarding. The research shows recycling and reselling are feasible options, emphasizes wiping data and removing accounts before disposal, and highlights information nudges to help people give old electronics a second life.
Nevada’s primary day proceeds with mostly mail voting and in-person turnout, but the secretary of state’s office briefly exposed testing results to The Nevada Independent’s server before polls closed. Officials say it was a secure testing environment and that results won’t be released until after the last voter in line at 7 p.m., as reporters continue to track turnout and issues across counties.