This week’s top media and technology stories highlight a decisive shift toward AI consolidation, automation and creator-led ecosystems, as major platforms streamline products, test advanced models and expand into new content, health and user engagement frontiers amid intensifying global competition.
OpenAI Plans ‘Superapp’ to Unify ChatGPT, Codex and Browser
OpenAI confirmed a Wall Street Journal report that it plans to merge its ChatGPT app, Codex coding platform and browser into a single desktop “superapp” designed to simplify the user experience. The move comes as the company seeks to streamline its product ecosystem and reduce fragmentation across multiple tools and stacks.
OpenAI President Greg Brockman will temporarily oversee the product overhaul and organisational changes, while Chief of Applications Fidji Simo will lead the sales team as the company prepares to market the unified app. In an internal note cited by the Journal, Simo acknowledged that spreading efforts across multiple apps had slowed progress and made it harder to meet quality standards.
Executives believe consolidation will improve efficiency and help OpenAI compete more effectively with rivals such as Anthropic. Earlier this year, the company launched a standalone desktop version of its Codex tool as part of its push into AI-powered code generation.
Mystery ‘Hunter Alpha’ Model Revealed as Xiaomi AI System
A powerful artificial intelligence model that surfaced anonymously on a developer platform has been revealed as a system developed by Chinese technology giant Xiaomi, ending speculation that startup DeepSeek was quietly testing its next-generation model, Reuters reported.
The model, named Hunter Alpha, appeared on OpenRouter on March 11 without attribution and was described as a “stealth model.” Its capabilities — including a reported one-trillion-parameter scale and a one-million-token context window — led observers to link it to the anticipated DeepSeek-V4 system.
Xiaomi’s AI team MiMo later confirmed that Hunter Alpha was an early internal test build of its MiMo-V2-Pro model, designed to act as the “brain” for AI agents capable of executing complex tasks with minimal human input. The model gained rapid adoption, surpassing one trillion tokens in usage.
The episode highlights growing competition in China’s AI sector, where stealth launches are increasingly used to gather unbiased feedback and test performance before formal releases.
Meta Expands AI Moderation, Reduces Reliance on Human Contractors
Meta announced plans to deploy more advanced artificial intelligence systems across its platforms, aiming to reduce reliance on third-party vendors for content moderation over the coming years. The company has already begun rolling out its Meta AI support assistant globally on Facebook and Instagram.
The assistant provides automated help for tasks such as password updates and profile settings, while broader AI systems are being introduced to detect violations including scams and illicit activity. Meta said human experts will remain responsible for training, oversight and high-impact decisions such as appeals and law enforcement referrals.
In early testing, Meta reported that its AI systems identified 5,000 scam attempts per day that human reviewers had missed, reduced impersonation reports by over 80%, and improved detection of certain violations while lowering error rates.
However, concerns persist around oversight, with reports of contractor lawsuits and instances of AI systems behaving unpredictably, underscoring the risks of increased automation.
Perplexity Launches iOS Browser with AI Assistant and Health Push
Perplexity launched its Comet browser app for iOS, marking a significant step in its expansion into mobile browsing and AI-powered search. The app integrates the company’s AI assistant directly into the browsing experience, enabling users to summarise content, answer questions and complete tasks such as making reservations.
Google has been set as the default search engine, chosen for its strength in handling fast, local queries. Comet also supports Perplexity’s Deep Research feature, which aggregates and summarises information from multiple web sources.
In parallel, the company introduced a health-focused initiative, including a new advisory board comprising physicians, researchers and health technology experts. The board will guide product decisions related to medical information, patient safety and clinical workflows.
Perplexity also expanded its Computer platform, enabling integration with health apps, wearable devices and medical records, signalling a broader push into personalised, data-driven services.
TikTok and Tubi Partner on Creator-Led Streaming Content
TikTok and streaming platform Tubi announced a “Creatorverse Incubator” partnership aimed at developing original shows created by TikTok talent for the free streaming service. The initiative will support both scripted and unscripted formats, with creators retaining significant creative direction.
TikTok will help identify participants and use its Spotlight promotional tool to build audience engagement around upcoming shows. Tubi will oversee production and distribution, positioning the programme as a bridge between short-form social content and long-form storytelling.
The move builds on Tubi’s growing catalogue of creator-led programming, which already includes more than 16,000 episodes from around 200 creators. The platform, which serves roughly 100 million monthly active users, has previously partnered with high-profile digital creators and studios.
Executives say the goal is to help creators expand beyond social media while bringing their existing audiences into the streaming ecosystem.
