Google Fixes Racial Discrimination Issues in Gemini AI Image Generation Tool

Translated data: Google's Gemini AI image generation tool has come under fire for potential racial discrimination risks in its diverse image outputs. In response, Google has quickly patched the tool to avoid accusations, emphasizing its commitment to diversity issues. Gemini has previously stumbled in this area, and Google needs to demonstrate its leadership in the field of AI development.

Listen to Garbled Audio and Get Hacked? Google Gemini Voice Assistant Exposes a Hidden Vulnerability, Hackers Use Special Notifications to Poison the AI

Smart homes and voice assistants have become new targets for hackers. SafeBreach revealed a hidden vulnerability in Google Gemini, allowing hackers to send malicious notifications via WhatsApp or SMS, tricking the voice assistant into performing unauthorized actions, potentially taking over smart homes or altering contact lists. This threat is named "Pseudo-Context Alignment".

Google Gemini Poisoned! New Vulnerability Exposed: Hackers Can Remotely Control Cars and Homes by Sending a Message with Hidden Information

Google Gemini Voice Assistant Exposes Serious Vulnerability: Hackers Can Send Special Notifications via WhatsApp or SMS to Bypass Security Defenses Using the "Pseudo-Context Alignment" Attack Method, Hiding Malicious Instructions in Non-English Text or Silent Links. Once Users Receive the Message, Gemini Can Be Jailbroken Without Their Awareness, Misjudging That the User Has Authorized It, Leading to Security Risks.

Lovable Deepens Strategic Partnership with Google, Migrating Infrastructure to Gemini and Google Cloud

Estonian AI startup Lovable (formerly GPT-Engineer.ai) has formed a strategic partnership with Google, migrating its infrastructure to Google Cloud and integrating the Gemini series of large models, replacing or supplementing the previously used Anthropic Claude 3.5 Sonnet. This move aims to provide a more scalable technical foundation for products that build end-to-end applications with natural language. The company was founded by Anton Osika, a former Mojang engineer.

Apple Siri Gets a Major Upgrade: A 1.2 Trillion-Parameter Google-Customized Model in Charge, Local Processing Speed Becomes Key

Apple is collaborating with Google to use a customized large language model with 1.2 trillion parameters as the core upgrade for the next generation of Siri. The scale of parameters in this model far exceeds current mainstream mobile models (such as Gemini 3.5 Flash with 300 billion parameters), drawing significant attention from the industry.

Apple AI Receives Epic Enhancement: Image Playground Will See a Significant Boost in OS27

Apple's previously highly publicized Apple Intelligence feature, particularly the image generation tool Image Playground, was humorously criticized by users as the worst AI function due to its poor output quality. However, the upcoming OS27 system will bring a turning point, with key upgrades including the adoption of a new foundational model and the introduction of Gemini computing power support, which is expected to significantly improve the performance of generative AI.