Microsoft makes Azure OpenAI Service generally available

Every enterprise wants in on generative AI, but corporate lawyers and security leads usually slam the brakes: What happens to our proprietary data? Who guarantees uptime? What about compliance?

Microsoft just handed those risk-averse executives an easy answer. The company officially announced the general availability of Azure OpenAI Service, throwing the enterprise doors wide open to the world's most capable AI models.

Enterprise armor for OpenAI's brains

First previewed back in November 2021, the service has now moved out of limited testing into prime time. Approved enterprise customers can tap into OpenAI's heavyweight roster, including GPT-3.5, the Codex code generator, and DALL-E 2 for image creation.

The real showstopper is still waiting in the wings. Microsoft confirmed that ChatGPT, which was trained and runs inference on Azure infrastructure, will be available through the enterprise service soon.

Getting in isn't completely friction-free, though. Microsoft is keeping its Limited Access framework in place. Developers must submit an application describing their specific use cases, and built-in content filters will actively monitor prompts and outputs for hateful or offensive material.

Real work, not just chatbot demos

Microsoft also shared what early adopters have actually been building, proving this isn't just about fun text generation.

Enterprise platform Moveworks integrated the models to scan corporate knowledge bases, spot missing documentation, and automatically draft new guides for internal HR and IT teams.

Al Jazeera Digital uses the service to speed up newsrooms with automated translation, summarization, and metadata tagging while strictly following editorial style guides.

Even accounting giant KPMG tapped the service for ESG tax compliance. Large multinational tax records are notoriously messy and buried deep inside legacy systems. KPMG uses Azure OpenAI to map data relationships, predict tax liabilities, and categorize payments across different jurisdictions.

Microsoft also reminded everyone that its own hit tools, GitHub Copilot, Power BI's natural-language formula generator, and the new Microsoft Designer, run on the exact same Azure foundation.

Why it matters

This move marks the transition of large language models from viral browser novelties into hardened corporate software.

For enterprise teams, it solves the single biggest bottleneck to adopting AI: legal clearance. By wrapping raw model capabilities in enterprise-grade security, uptime guarantees, and strict data privacy, Microsoft lets tech leads build without betting the company's data privacy.

For the broader cloud wars, this puts massive heat on AWS and Google Cloud. Microsoft now has an exclusive, enterprise-ready pipeline to the most hyped AI systems on the planet. Many enterprise accounts could easily shift parts of their cloud spending to Azure purely to deploy compliant GPT workloads.

That said, Microsoft's gating process could still frustrate smaller developers. The strict application review and content guardrails are smart risk-management moves for corporate clients, but they could push fast-moving startups back toward lighter, unvetted alternatives.