Microsoft at the moment at its annual Construct convention launched a number of updates to Azure AI, the corporate’s cloud-based platform for constructing and operating AI functions. Azure AI competes with related choices from rival cloud suppliers comparable to AWS, Google, and IBM.
The updates embody the addition of recent governance options, new massive language fashions (LLMs), and Azure AI Search enhancements. Microsoft additionally introduced that it’s making Azure AI Studio usually accessible.
Azure AI Studio, a generative AI software growth toolkit that competes with the likes of Amazon Bedrock and Google Vertex AI Studio, was launched in a preview in November of final 12 months.
In distinction to Microsoft’s Copilot Studio providing, which is a low-code device for customizing chatbots, Azure AI Studio is aimed toward skilled builders, permitting them to decide on generative AI fashions and floor them with retrieval augmented technology (RAG) utilizing vector embeddings, vector search, and their very own knowledge sources.
Azure AI Studio will also be used to fine-tune fashions and create AI-powered copilots or brokers.
New fashions added to Azure AI
As a part of the updates to Azure AI, Microsoft is including new fashions to the mannequin catalog inside Azure AI Studio, bringing the variety of fashions accessible to greater than 1,600.
The brand new fashions embody OpenAI’s GPT-4o, showcased this week. Earlier in Might, Microsoft enabled GPT-4 Turbo with Imaginative and prescient by way of Azure OpenAI Service. “With these new fashions builders can construct apps with inputs and outputs that span throughout textual content, photographs, and extra,” the corporate stated in an announcement.
Different fashions which have been added by way of Azure AI’s Fashions-as-a-Service (MaaS) providing embody TimeGen-1 from Nixtla and Core42 JAIS, which are actually accessible in preview. Fashions from AI21, Bria AI, Gretel Labs, NTT Information, Stability AI, and Cohere Rerank are anticipated to be added quickly, Microsoft stated.
Additional, Microsoft is updating its Phi-3 household of small language fashions (SLMs) with the addition of Phi-3-vision, a brand new multimodal mannequin that’s anticipated to change into accessible in preview.
In April, Microsoft had launched three Phi-3 fashions—the three.8-billion-parameter Phi-3 Mini, the 7-billion-parameter Phi-3 Small, and the 14-billion-parameter Phi-3 Medium—to assist resource-constrained environments for on-device, edge, and offline inferencing and be cheaper for enterprises.
Microsoft’s Phi-3 builds on Phi-2, which may perceive 2.7 billion parameters whereas outperforming massive language fashions as much as 25 instances bigger, Microsoft stated on the time of the launch. Phi-3 Mini is presently usually accessible as a part of Azure AI’s Fashions-as-a-Service providing.
Different modules of Azure AI have been additionally up to date, together with Azure AI Speech, which now consists of options comparable to speech analytics and common translation to assist builders construct functions to be used circumstances requiring audio enter and output. The brand new options can be found in preview.
Again in April, Microsoft had up to date its Azure AI Search service to extend storage capability and vector index dimension at no further value, a transfer it stated will make it extra economical for enterprises to run generative AI-based functions.
Azure AI will get new governance, security options
At Construct 2024 Microsoft additionally launched new governance and security options for Azure AI, with the corporate updating its mannequin output monitoring system, Azure AI Content material Security.
The brand new function, named Customized Classes, is presently in preview and can enable builders to create customized filters for particular content material filtering wants. “This new function additionally features a speedy choice, enabling you to deploy new customized filters inside an hour to guard in opposition to rising threats and incidents,” Microsoft stated.
Different governance options added to Azure AI Studio and Azure OpenAI Service embody Immediate Shields and Groundedness Detection, each of that are in preview.
Whereas Immediate Shields mitigate each oblique and jailbreak immediate injection assaults on LLMs, Groundedness Detection checks generative AI functions for ungrounded outputs or hallucinations in generated responses.
Microsoft stated that it presently has 20 accountable AI instruments with greater than 90 options throughout its choices and companies.
With a view to safe generative AI functions, Microsoft stated that it was integrating Microsoft Defender for Cloud throughout all of its AI companies. “Risk safety for AI workloads in Defender for Cloud leverages a local integration with Azure AI Content material Security to allow safety groups to watch their Azure OpenAl functions for direct and in-direct immediate injection assaults, delicate knowledge leaks, and different threats to allow them to rapidly examine and reply,” the corporate stated.
With a view to bear down on safety additional, enterprise builders may also combine Microsoft Purview into their developed functions and copilots with the assistance of APIs, based on Jessica Hawk, company vice chairman of knowledge, AI, and digital functions at Microsoft.
It will assist builders and copilot clients to find knowledge dangers in AI interactions, shield delicate knowledge with encryption, and govern AI actions, Hawk added.
These capabilities can be found for Copilot Studio in public preview and can be accessible in public preview for Azure AI Studio in July by way of the Purview SDK.
Different safety updates embody integration of what Microsoft calls “hidden layers safety scanning” into Azure AI Studio to scan each mannequin for malware.
One other function, referred to as Facial Liveness, has been added to the Azure AI Imaginative and prescient Face API. “Home windows Whats up for Enterprise makes use of Facial Liveness as a key ingredient in multi-factor authentication (MFA) to stop spoofing assaults,” Hawk defined.
Copyright © 2024 IDG Communications, Inc.
