Microsoft prepares itself to host OpenAI's GPT-5 and other models, as AI competition increases

Microsoft is gearing up to host OpenAI’s next big AI models, GPT-4.5 and GPT-5, as it expands server capacity in anticipation of their rollout. While OpenAI CEO Sam Altman recently hinted at the launch of GPT-4.5 in the coming weeks, Microsoft is reportedly expecting to integrate the model as early as next week.

Codenamed Orion, GPT-4.5 is positioned as OpenAI’s last non-chain-of-thought model, promising a noticeable improvement over GPT-4. However, the real breakthrough is expected with GPT-5, which could launch by late May. This timeline aligns with Altman’s previous statements that the next-generation model would arrive within months. Initially planned for a late 2024 release, GPT-4.5 was pushed to early 2025, paving the way for GPT-5’s much-anticipated debut.

What’s New in GPT-5?

GPT-5 is expected to be a major milestone for OpenAI, combining multiple AI technologies into one seamless system. A key addition will be the integration of OpenAI’s o3 reasoning model, which was first teased in December during the company’s holiday announcements. Unlike its predecessor, o3-mini, this model will not be released separately but will instead become a core part of GPT-5, further advancing OpenAI’s goal of developing artificial general intelligence (AGI).

Another significant change will be the way users interact with ChatGPT. OpenAI is working to remove the need for users to select AI models for different tasks manually. Altman recently acknowledged that the team dislikes the current model-selection process as much as users do, and GPT-5 will aim to provide a “magic unified intelligence” experience where the AI adapts dynamically without requiring manual input.

More from Tech

Microsoft unveils new chip that could make quantum computing go mainstream in just few years Under pressure from DeepSeek, OpenAI to roll out bunch of new features for all ChatGPT users

Microsoft’s AI strategy and the GPT-5 rollout

Microsoft is expected to launch GPT-5 integration around its Build developer conference, scheduled for May 19. The timing aligns with Google I/O, which will take place the same week, setting up a direct AI showdown between the two tech giants.

Last year, OpenAI’s sudden GPT-4o launch disrupted Microsoft’s own AI services, as it undercut some of the company’s paid Azure AI offerings. However, this time, Microsoft is fully prepared, with planned updates to its AI assistant, Copilot, set to follow shortly after OpenAI’s latest releases. In recent months, Copilot has seen major updates, including a revamped interface and the removal of manual model-selection options.

Microsoft is also working on an AI agent similar to OpenAI’s Operator, designed to automate web-based tasks by interacting with graphical interfaces. This feature functions like an autopilot for the web, allowing users to complete actions without direct input, making AI-powered automation more accessible.

Editor’s Picks

1

Amazon, Microsoft, Alphabet, and Meta to spend well over $300 billion on AI in 2025, fearing China 2

Microsoft unveils new chip that could make quantum computing go mainstream in just few years

AI showdown at Microsoft Build and Google I/O

With Microsoft Build and Google I/O happening simultaneously, both companies are expected to showcase their latest AI advancements in a bid to outshine one another. Microsoft has been focusing heavily on cost reduction and efficiency in its AI services, making them more attractive for enterprise customers.

As OpenAI’s most advanced model yet, GPT-5 is set to bring significant improvements, and Microsoft is ensuring it’s ready to capitalise on the launch. With AI developments moving at lightning speed, the competition between Microsoft, OpenAI, and Google is about to reach a new level.