3 AI Agent Updates: IBM, Dolphin AI and Ajax|AI News

THREE AI AGENTS DEVELOPMENTS CHANGING THE AI LANDSCAPE

AI agents are moving rapidly from experimental demonstrations into practical products that can perform useful tasks for businesses, creators and individual users. Three recent developments highlight three different directions for the future of agentic AI: sovereign AI infrastructure in India, automated video production, and private AI agents running locally.

IBM and Yotta bring sovereign AI agents to India

IBM and Yotta Data Services announced the general availability of a sovereign agentic AI platform for Indian organizations. The platform combines IBM watsonx Orchestrate with Yotta's Shakti Cloud infrastructure.

The key idea behind sovereign AI is control. Organizations can deploy AI agents while keeping data, inference, operations and governance controls within India. This can be especially important for businesses and institutions that handle sensitive information and have requirements around where their data is processed.

The platform is designed for business workflows such as security operations, document processing and HR automation. Yotta's infrastructure includes cloud regions in Panvel and Greater Noida.

This development shows that AI adoption is increasingly becoming an infrastructure and governance question—not simply a question of which AI model is most powerful.

Eloelo launches Dolphin AI for video production

Indian startup Eloelo has also entered the agentic AI space with Dolphin AI, a video-production platform aimed at creators, brands, filmmakers and studios.

Dolphin AI is designed to take a script through several stages of production, including planning, shot generation, editing and enhancement. One of its notable goals is maintaining continuity between shots.

AI-generated videos can often suffer from inconsistent characters, clothing, voices, props or visual styles when scenes are created separately. Dolphin AI is designed to help maintain these elements across multiple shots, making it more suitable for structured storytelling and longer productions.

The workflow can be summarized as:

Script → Shot planning → Scene generation → Editing → Enhancement → Final video

PewDiePie's Ajax focuses on local AI

A third development comes from creator PewDiePie, whose Ajax project is designed as a local AI agent for the Odysseus workspace.

Ajax is described as a fine-tuned Qwen 3.5 9B model intended to support tasks such as web browsing, email and calendar management while operating locally.

Local AI can offer an important advantage: greater control over where information is processed. Instead of sending every task to a large cloud-based model, smaller specialized models can potentially perform useful tasks directly on a user's own hardware.

These three developments reveal a broader trend in artificial intelligence. AI agents are no longer being developed only as general-purpose chatbots. They are becoming specialized systems for enterprise automation, creative production and private personal computing.

The future of AI may therefore involve many different agents working in different environments: sovereign agents for