Retter Reboot #50

OpenAI Responds to Google With GPT-5.2 After “Code Red” Memo

OpenAI has released GPT-5.2, positioning it as a direct response to Google’s internal “code red” push around AI competition. The new model focuses on stronger reasoning, faster responses, and improved coding and agentic capabilities, aiming to widen the gap with rival models from Google and others. The launch underscores how intensifying competition between OpenAI and Google is accelerating model releases, with both companies racing to claim leadership in advanced, general-purpose AI systems.

Source: Techcrunch

Google Launches Its Deepest AI Research Agent as OpenAI Releases GPT-5.2

Google has unveiled its most advanced AI research agent to date, designed to autonomously plan, execute, and synthesize complex research tasks across the web and internal tools. The agent can break down long-term goals, gather sources, run experiments, and produce structured outputs with minimal human input. The launch came the same day OpenAI released GPT-5.2, highlighting the escalating arms race between the two companies as they push toward more capable, agentic AI systems focused on reasoning, research, and automation.

Source: Techcrunch

Google Debuts Disco, a Gemini-Powered Tool That Builds Web Apps From Browser Tabs

Google has introduced Disco, a new Gemini-powered tool that can turn open browser tabs into working web apps. Disco analyzes the content, structure, and intent of tabs, then automatically generates app logic, UI components, and workflows, allowing users to prototype or build apps without starting from scratch. The launch reinforces Google’s push to make AI-assisted app creation more accessible, blurring the line between browsing, prototyping, and development.

Source: Techcrunch

Google’s AI Virtual Try-On Now Works With Just a Selfie

Google has updated its AI-powered virtual try-on feature so users can see how clothes might look on them using a single selfie. The system maps garments onto the user’s body and adjusts fit and appearance based on pose and proportions, without needing full-body scans or measurements. The feature aims to make online fashion shopping more intuitive, reducing uncertainty around fit and helping users visualize outfits more realistically before buying.

Source: Techcrunch

Disney Signs Deal With OpenAI to Let Sora Generate Videos Featuring Its Characters

Disney has reached a new agreement with OpenAI that allows Sora to generate AI videos featuring Disney-owned characters under controlled conditions. The partnership focuses on licensed use cases for marketing, storytelling, and creative experimentation, with safeguards to protect Disney’s intellectual property and brand integrity. The deal marks a major shift for Hollywood, showing how large media companies are beginning to embrace generative video AI rather than fight it, as long as usage is licensed and tightly governed.

Source: Techcrunch

RSL Launches a New Licensing Spec for AI Training and Publishing

RSL has introduced a new standardized licensing specification designed to clarify how content can be used for AI training and AI-generated publishing. The spec aims to give publishers and platforms a clear, machine-readable way to signal what AI systems are allowed to do with their content, including training, summarization, and reuse. The move is part of a broader push to create industry-wide rules for AI content usage, as tensions grow between AI companies and publishers over data rights and compensation.

Source: The Verge

DeepSeek Allegedly Acquires Banned Nvidia Blackwell Chips Amid Smuggling Bust

Chinese AI startup DeepSeek is reportedly linked to the acquisition of restricted Nvidia Blackwell GPUs, as U.S. authorities uncovered a $160 million chip-smuggling operation. The U.S. Department of Justice says the network illegally routed advanced AI chips to China, bypassing export controls meant to limit access to cutting-edge compute. The case highlights how AI demand is straining global export regulations, with advanced chips becoming a critical and tightly controlled resource in the geopolitical race for AI dominance.

Source: WinBuzzer

OpenAI, Anthropic, and Block Join Linux Foundation Effort to Standardize AI Agents

OpenAI, Anthropic, and Block have joined a new Linux Foundation initiative aimed at creating open standards for AI agents. The effort focuses on defining shared protocols, interfaces, and governance models so AI agents can work across platforms, tools, and ecosystems more reliably. The move signals a push toward interoperability in the agentic AI era, as companies seek common foundations to accelerate adoption while avoiding fragmented, proprietary systems.

Source: https://techcrunch.com/2025/12/09/openai-anthropic-and-block-join-new-linux-foundation-effort-to-standardize-the-ai-agent-era/

Google Plans to Launch AI-Powered Smart Glasses in 2026

Google is preparing to release AI-powered smart glasses in 2026, aiming to re-enter the wearables market with a stronger focus on Gemini-driven intelligence. The glasses are expected to offer real-time assistance such as navigation, translation, visual search, and contextual notifications, tightly integrated with Android and Google services. The move reflects renewed confidence in AI-first wearables, as advances in models, on-device processing, and hardware design make smart glasses more practical and socially acceptable than earlier attempts.

Source: CNBC

EU Opens Antitrust Probe Into Google’s AI Search Use of Publisher Content

The European Union has launched an antitrust investigation into Google, focusing on how its AI-powered search results use content from publishers. Regulators are examining whether Google’s AI summaries unfairly reuse or replace original articles, potentially diverting traffic and revenue away from news organizations. The probe adds pressure on Google as publishers argue that AI search features undermine their business models, intensifying the broader debate over fair compensation and content rights in the age of generative AI.

Source: WinBuzzer