According to reports monitored by Beating, Grok's web version is rolling out Imagine Agent Mode (beta), a new agent-based tool built on xAI's existing image and video generation capabilities. The interface replaces the traditional chat window with an infinite canvas and includes four preset

According to Bloomberg, Apple plans to add a Siri-branded AI mode to the iPhone camera app in iOS 27, moving its Visual Intelligence tool from the Camera Control button into the main camera interface. The mode would appear alongside Photo and Video options, allowing users to point the camera at

Gate News message, April 29 — EVAA announced the launch of E-Mode (Efficiency Mode) on April 29, marking a new era of capital efficiency in the TON ecosystem. Based on TON's current interest rate of approximately 24%, E-Mode can increase strategy annual yields to as high as 41%, significantly

Gate News message, April 29 — OpenAI CEO Sam Altman announced a redesigned Codex interface on X today, introducing two distinct modes for users. "Excelmogging" targets everyday office tasks with a simplified interface and the tagline "Same tools, simpler interface," featuring example tasks like

Gate News message, April 25 — OpenClaw, an open-source AI agent framework, released v2026.4.23 on April 23, introducing updates across image generation, sub-agent mechanisms, and security hardening. Image generation enhancements allow gpt-image-2 to be called directly via Codex OAuth without

Gate News message, April 23 — GoPlus AI security component AgentGuard has officially released its Checkup feature, designed to address security blind spots in AI agents' autonomous decision-making and asset control, elevating AI agent security from "fragmented interception" to "systematic

Crypto Fear & Greed Index fell to 32 from 33, signaling continued panic; the 0-100 scale marks fear below 50 and greed above 50. Abstract: The Crypto Fear & Greed Index dropped to 32 on April 22, signaling ongoing panic after a decline from 33 the previous day. The index operates on a 0-100 scale, with readings below 50 indicating fear and readings above 50 indicating greed.

Anthropic在3月31日意外泄露了Claude Code的51.2万行源代码，引发安全风险和对市场估值的担忧。此次泄露后，代码迅速传播至GitHub，并引起DMCA删除。泄露内容揭示了内部功能和未发布细节，暴露了企业面临的安全风险。事件强调AI企业需加强代码安全与权限管理，同时可能影响Anthropic的上市前景。

蚂蚁集团工程师陈成逆向Claude Code 2.1.81源码，揭示其Auto Mode决策机制。这一机制通过四层决策流水线进行工具调用的判断，只有在前三层都无法判断时，才调用独立的AI分类器进行安全审查。分类器设计专注于风险监控，设有熔断机制及行为提示词注入频率控制。

Anthropic推出Claude Code的新权限模式auto mode，通过预先审查操作安全性，在逐项审批与完全跳过权限之间提供更灵活选择。安全操作自动放行，危险操作被拦截并需用户确认。该模式降低风险但不完全消除，建议在隔离环境中使用，现已向部分用户开放。