Microsoft cautions that rapid deployment of workplace AI assistants can turn them into insider threats, calling the risk a "double agent." The company’s Cyber Pulse report explains how attackers can manipulate an agent’s access or feed it malicious input, using its legitimate privileges to cause damage inside an organization. Microsoft urges firms to treat AI agents as a new class of digital identity, apply Zero Trust principles, enforce least‑privilege access, and maintain centralized visibility to prevent memory‑poisoning attacks and other forms of tampering. Leia mais →

OpenAI has rolled out a global age‑prediction system for ChatGPT to automatically apply a teen‑mode experience to users it believes are under 18. The model relies on behavioral cues, account history, usage patterns, and language analysis, and defaults to caution when uncertain. Several adult subscribers report being mistakenly routed to teen mode, facing content restrictions and being asked to verify their age through a third‑party tool that may request official ID or a selfie video. Users criticize the invasive verification process and raise privacy concerns, while OpenAI says the data is deleted after verification and promises ongoing refinements. Leia mais →

OpenAI announced a global rollout of an age prediction system for ChatGPT accounts. The model evaluates behavioral and account-level signals to estimate whether a user is a minor, and users flagged as underage must verify their age through a selfie on the Persona age verification platform. The move follows criticism of AI firms for adding safety measures after incidents, including a wrongful‑death lawsuit linked to a teen’s use of ChatGPT. OpenAI also plans an "adult mode" for NSFW content, prompting concerns that underage users may try to bypass the new protections, similar to challenges seen on platforms like Roblox. Leia mais →

A growing number of AI applications let users discover celebrity look‑alikes by uploading personal photos. While ChatGPT itself cannot perform facial recognition, users can access specialized GPTs that accept images and generate match suggestions. The process involves selecting clear, front‑facing photos without accessories or filters, uploading them to a dedicated GPT, and receiving a list of possible celebrity matches. Users have reported mixed results—some tools suggest familiar faces, while others produce surprising or humorous outcomes. Privacy concerns remain, as the technology relies on image analysis and pattern matching. Leia mais →

Los Angeles Comic Con is planning an artificial‑intelligence‑driven hologram of Stan Lee, aiming to let fans interact with a digital version of the Marvel legend. The concept builds on the convention’s earlier use of a volumetric hologram for Boffo the Bear, which relied on live motion capture. This new venture will combine the holographic platform created by Proto Hologram with an AI model supplied by Hyperreal, a firm that focuses on digital identity ownership. Organizers hope the avatar will answer questions about the Marvel universe while staying true to Lee’s own words. Leia mais →