Microsoft Copilot Gains Agentic AI Capabilities with Actions and Deep Research

Microsoft Copilot Enhances AI Capabilities with Actions and Deep Research

Microsoft has significantly advanced its Copilot AI stack with the introduction of two key features: Actions and Deep Research. These updates position Copilot as a more capable agentic AI, designed to perform complex, multi-step web-based tasks autonomously or semi-autonomously on behalf of users.

Agentic AI and Copilot's New Frontier

Agentic AI refers to AI tools that can execute a series of actions to achieve a goal. Copilot's new Actions feature allows it to handle tasks such as booking event tickets, making dinner reservations, or sending gifts, effectively checking items off a user's to-do list. This moves beyond traditional AI chatbots that primarily provide information based on input, enabling Copilot to perform real-world actions through natural language commands.

How Copilot Actions Work

Users can instruct Copilot to perform tasks like finding nearby restaurants that are open late and then booking a table. Copilot achieves this by browsing the web, filling in necessary details, and seeking user input when required. This streamlines processes that would typically involve manual website navigation and interaction.

Partnerships for Enhanced Functionality

To facilitate these actions, Microsoft has partnered with several prominent companies, including 1-800-Flowers.com, Booking.com, Expedia, Kayak, OpenTable, Priceline, Tripadvisor, Skyscanner, Viator, and Vrbo. These collaborations enable Copilot to execute tasks across a wide range of services.

Competitive Landscape of AI Agents

Microsoft is not the first to introduce such capabilities. OpenAI's Operator has been available for some time, offering similar functionalities. Amazon is also developing a comparable system called Nova Act, built on its Nova AI model, which aims to automate tasks like ordering food and restocking supplies. This technology is already integrated into Amazon's new Alexa+ assistant. Furthermore, Amazon is testing an AI agent in its mobile app that can purchase items from other retailers' websites. Even the Opera browser is developing its own AI operator to handle basic web-based tasks.

Deep Research Feature

While the article primarily focuses on Actions, it also mentions the Deep Research feature, suggesting an enhanced capability for Copilot to gather and synthesize information more thoroughly. This would complement the action-oriented features by providing a stronger foundation of knowledge for decision-making.

Impact and Future of AI Assistants

The introduction of Actions and Deep Research signifies a significant step towards more integrated and functional AI assistants. By enabling AI to perform real-world tasks, Microsoft is pushing the boundaries of what AI can do, making digital interactions more seamless and efficient. This trend towards agentic AI is likely to reshape how users interact with technology, offering greater convenience and productivity.

Author and Related Content

The article is authored by Nadeem Sarwar, a tech and science journalist. The content is categorized under Artificial Intelligence, Digital Trends Videos, Microsoft, Microsoft Copilot, and Tech News. Related articles cover topics such as AI's role in music, Apple's AI strategy, AI-powered cameras, and comparisons between AI search engines like Gemini Advanced, ChatGPT, and Copilot Pro.

Image and Visuals

The article includes an image of Microsoft Copilot Actions, showcasing a search with the feature. It also references other visual content related to NASA's VR training, Dyson's growing systems, Wimbledon's robot judges, and various AI-related news items.

Conclusion

Microsoft's latest updates to Copilot, particularly the Actions and Deep Research features, represent a significant leap in the development of agentic AI. These advancements aim to provide users with a more powerful and versatile AI companion capable of handling complex tasks, setting a new benchmark in the competitive AI landscape.