# Microsoft AI

Microsoft AI + Azure OpenAI.

Status: provisional
Last verified: 2026-08-27T21:50:18.000Z

## Evidence

- (single_source) Forrester interviewed 10 decision-makers at five organizations and surveyed 154 other decision-makers and AI leaders across the U.S. and Europe with experience using Microsoft Foundry.
  - supports https://azure.microsoft.com/en-us/blog/the-economics-of-enterprise-ai-what-the-forrester-tei-study-reveals-about-microsoft-foundry
    > When Forrester modeled the economics of enterprise AI with Microsoft Foundry, the biggest driver behind the 327% ROI over three years 1 was surprising: developer productivity, worth $15.7 million over the same period.
...
35%. Teams using Foundry to develop AI apps and agents saw payback in as few as six months and with benefits accelerating year over year1.
...
Forrester interviewed 10 decision-makers at five organizations and surveyed 154 other decision-makers and AI leaders across the U.S. and Europe with experience using Microsoft Foundry. They modeled a composite enterprise with $10 billion revenue, 25,000 employees, and 100 technical staff using Foundry. To model conservative estimates, benefits were adjusted downward and costs upward; the results reflect the composite enterprise.

- (single_source) Forrester modeled a composite enterprise with $10 billion revenue, 25,000 employees, and 100 technical staff using Foundry.
  - supports https://azure.microsoft.com/en-us/blog/the-economics-of-enterprise-ai-what-the-forrester-tei-study-reveals-about-microsoft-foundry
    > When Forrester modeled the economics of enterprise AI with Microsoft Foundry, the biggest driver behind the 327% ROI over three years 1 was surprising: developer productivity, worth $15.7 million over the same period.
...
35%. Teams using Foundry to develop AI apps and agents saw payback in as few as six months and with benefits accelerating year over year1.
...
Forrester interviewed 10 decision-makers at five organizations and surveyed 154 other decision-makers and AI leaders across the U.S. and Europe with experience using Microsoft Foundry. They modeled a composite enterprise with $10 billion revenue, 25,000 employees, and 100 technical staff using Foundry. To model conservative estimates, benefits were adjusted downward and costs upward; the results reflect the composite enterprise.

- (single_source) Forrester modeled a 327% ROI over three years for Microsoft Foundry, with developer productivity being the biggest driver, worth $15.7 million over the same period.
  - supports https://azure.microsoft.com/en-us/blog/the-economics-of-enterprise-ai-what-the-forrester-tei-study-reveals-about-microsoft-foundry
    > When Forrester modeled the economics of enterprise AI with Microsoft Foundry, the biggest driver behind the 327% ROI over three years 1 was surprising: developer productivity, worth $15.7 million over the same period.
...
35%. Teams using Foundry to develop AI apps and agents saw payback in as few as six months and with benefits accelerating year over year1.
...
Forrester interviewed 10 decision-makers at five organizations and surveyed 154 other decision-makers and AI leaders across the U.S. and Europe with experience using Microsoft Foundry. They modeled a composite enterprise with $10 billion revenue, 25,000 employees, and 100 technical staff using Foundry. To model conservative estimates, benefits were adjusted downward and costs upward; the results reflect the composite enterprise.

- (single_source) PersonaTeaming Playground is a user-facing interface that enables red-teamers to author their own personas and collaborate with AI to mutate and refine prompts.
  - supports https://microsoft.com/en-us/research/publication/personateaming-supporting-persona-driven-red-teaming-for-generative-ai
    > . In this work, we explore persona-driven red-teaming to advance both automated red-teaming and human-AI collaboration. We first develop PersonaTeaming Workflow, which incorporates personas into the adversarial prompt generation process to explore a wider spectrum of adversarial strategies. Compared to RainbowPlus, a state-of-the-art automated red-teaming method, PersonaTeaming Workflow achieves higher attack success rates while maintaining prompt diversity. However, since automated personas only approximate real human perspectives, we further instantiate PersonaTeaming Workflow as PersonaTeaming Playground, a user-facing interface that enables red-teamers to author their own personas and collaborate with AI to mutate and refine prompts. In a user study with 11 industry practitioners, we

- (single_source) Compared to RainbowPlus, a state-of-the-art automated red-teaming method, PersonaTeaming Workflow achieves higher attack success rates while maintaining prompt diversity.
  - supports https://microsoft.com/en-us/research/publication/personateaming-supporting-persona-driven-red-teaming-for-generative-ai
    > . In this work, we explore persona-driven red-teaming to advance both automated red-teaming and human-AI collaboration. We first develop PersonaTeaming Workflow, which incorporates personas into the adversarial prompt generation process to explore a wider spectrum of adversarial strategies. Compared to RainbowPlus, a state-of-the-art automated red-teaming method, PersonaTeaming Workflow achieves higher attack success rates while maintaining prompt diversity. However, since automated personas only approximate real human perspectives, we further instantiate PersonaTeaming Workflow as PersonaTeaming Playground, a user-facing interface that enables red-teamers to author their own personas and collaborate with AI to mutate and refine prompts. In a user study with 11 industry practitioners, we

- (single_source) PersonaTeaming Workflow incorporates personas into the adversarial prompt generation process to explore a wider spectrum of adversarial strategies.
  - supports https://microsoft.com/en-us/research/publication/personateaming-supporting-persona-driven-red-teaming-for-generative-ai
    > . In this work, we explore persona-driven red-teaming to advance both automated red-teaming and human-AI collaboration. We first develop PersonaTeaming Workflow, which incorporates personas into the adversarial prompt generation process to explore a wider spectrum of adversarial strategies. Compared to RainbowPlus, a state-of-the-art automated red-teaming method, PersonaTeaming Workflow achieves higher attack success rates while maintaining prompt diversity. However, since automated personas only approximate real human perspectives, we further instantiate PersonaTeaming Workflow as PersonaTeaming Playground, a user-facing interface that enables red-teamers to author their own personas and collaborate with AI to mutate and refine prompts. In a user study with 11 industry practitioners, we

- (single_source) PersonaTeaming Playground was evaluated in a user study with 11 industry practitioners.
  - supports https://microsoft.com/en-us/research/publication/personateaming-supporting-persona-driven-red-teaming-for-generative-ai
    > . In this work, we explore persona-driven red-teaming to advance both automated red-teaming and human-AI collaboration. We first develop PersonaTeaming Workflow, which incorporates personas into the adversarial prompt generation process to explore a wider spectrum of adversarial strategies. Compared to RainbowPlus, a state-of-the-art automated red-teaming method, PersonaTeaming Workflow achieves higher attack success rates while maintaining prompt diversity. However, since automated personas only approximate real human perspectives, we further instantiate PersonaTeaming Workflow as PersonaTeaming Playground, a user-facing interface that enables red-teamers to author their own personas and collaborate with AI to mutate and refine prompts. In a user study with 11 industry practitioners, we

- (single_source) Teams using Microsoft Foundry to develop AI apps and agents saw payback in as few as six months, with benefits accelerating year over year.
  - supports https://azure.microsoft.com/en-us/blog/the-economics-of-enterprise-ai-what-the-forrester-tei-study-reveals-about-microsoft-foundry
    > When Forrester modeled the economics of enterprise AI with Microsoft Foundry, the biggest driver behind the 327% ROI over three years 1 was surprising: developer productivity, worth $15.7 million over the same period.
...
35%. Teams using Foundry to develop AI apps and agents saw payback in as few as six months and with benefits accelerating year over year1.
...
Forrester interviewed 10 decision-makers at five organizations and surveyed 154 other decision-makers and AI leaders across the U.S. and Europe with experience using Microsoft Foundry. They modeled a composite enterprise with $10 billion revenue, 25,000 employees, and 100 technical staff using Foundry. To model conservative estimates, benefits were adjusted downward and costs upward; the results reflect the composite enterprise.

- (single_source) To model conservative estimates in the Forrester TEI study, benefits were adjusted downward and costs upward.
  - supports https://azure.microsoft.com/en-us/blog/the-economics-of-enterprise-ai-what-the-forrester-tei-study-reveals-about-microsoft-foundry
    > When Forrester modeled the economics of enterprise AI with Microsoft Foundry, the biggest driver behind the 327% ROI over three years 1 was surprising: developer productivity, worth $15.7 million over the same period.
...
35%. Teams using Foundry to develop AI apps and agents saw payback in as few as six months and with benefits accelerating year over year1.
...
Forrester interviewed 10 decision-makers at five organizations and surveyed 154 other decision-makers and AI leaders across the U.S. and Europe with experience using Microsoft Foundry. They modeled a composite enterprise with $10 billion revenue, 25,000 employees, and 100 technical staff using Foundry. To model conservative estimates, benefits were adjusted downward and costs upward; the results reflect the composite enterprise.

- (single_source) Azure pricing is calculated based on US dollars and converted using London closing spot rates captured in the two business days prior to the last business day of the previous month end.
  - supports https://azure.microsoft.com/en-us/pricing/details/ai-foundry-models/llama
    > as actual price quotes. Actual pricing may vary depending on the type of agreement entered with Microsoft, date of purchase, and the currency exchange rate. Prices are calculated based on US dollars and converted using London closing spot rates that are captured in the two business days prior to the last business day of the previous month end. If the two business days prior to the end of the month fall on a bank holiday in major markets, the rate setting day is generally the day immediately preceding the two business days. This rate applies to all transactions during the upcoming month. Sign in to the Azure pricing calculator to see pricing based on your current program/offer with Microsoft. Contact an Azure sales specialist for more information on pricing or to request a price quote. See

- (single_source) Azure OpenAI Service delivers enterprise-ready generative AI featuring powerful models from OpenAI, enabling organizations to innovate with text, audio, and vision capabilities.
  - supports https://azure.microsoft.com/en-us/pricing/details/azure-openai
    > Azure OpenAI Service - Pricing | Microsoft Azure
...
# Azure OpenAI pricing
...
Azure OpenAI Service delivers enterprise-ready generative AI featuring powerful models from OpenAI, enabling organizations to innovate with text, audio, and vision capabilities. Beyond the cutting-edge models, companies choose Azure OpenAI Service for built-in data privacy, regional/area/global flexibility, and seamless integration into the Azure ecosystem including Fabric, Cosmos DB and Azure AI Search. Companies of all sizes can confidently scale AI solutions to enhance customer experience, automate workflows, and unlock creative potential, driving measurable impact and competitive differentiation. To help customers in the journey, we offer pricing and cost management solutions to meet your needs. including:

- (single_source) Companies choose Azure OpenAI Service for built-in data privacy, regional/area/global flexibility, and seamless integration into the Azure ecosystem including Fabric, Cosmos DB and Azure AI Search.
  - supports https://azure.microsoft.com/en-us/pricing/details/azure-openai
    > Azure OpenAI Service - Pricing | Microsoft Azure
...
# Azure OpenAI pricing
...
Azure OpenAI Service delivers enterprise-ready generative AI featuring powerful models from OpenAI, enabling organizations to innovate with text, audio, and vision capabilities. Beyond the cutting-edge models, companies choose Azure OpenAI Service for built-in data privacy, regional/area/global flexibility, and seamless integration into the Azure ecosystem including Fabric, Cosmos DB and Azure AI Search. Companies of all sizes can confidently scale AI solutions to enhance customer experience, automate workflows, and unlock creative potential, driving measurable impact and competitive differentiation. To help customers in the journey, we offer pricing and cost management solutions to meet your needs. including:

- (single_source) Microsoft Research argues that many attack success rate (ASR) comparison conclusions in AI red teaming are founded on apples-to-oranges comparisons or low-validity measurements.
  - supports https://microsoft.com/en-us/research/publication/comparison-requires-valid-measurement-rethinking-attack-success-rate-comparisons-in-ai-red-teaming
    > rate (ASR) comparisons. We show, through conceptual, theoretical, and empirical contributions, that many conclusions are founded on apples-to-oranges comparisons or low-validity measurements. Our arguments are grounded in asking a simple question: (When) can attack success rates be meaningfully compared? To answer this question, we draw on ideas from social science measurement theory and inferential statistics, which, taken together, provide a conceptual grounding for understanding when numerical values obtained by quantification of system attributes can be meaningfully compared. Through this lens, we articulate sufficient conditions under which ASRs can be meaningfully compared. Using jailbreaking as a running example, we provide examples and extensive discussion of apples-to-oranges ASR

- (single_source) The Microsoft Research publication on ASR comparisons draws on ideas from social science measurement theory and inferential statistics to articulate sufficient conditions under which ASRs can be meaningfully compared.
  - supports https://microsoft.com/en-us/research/publication/comparison-requires-valid-measurement-rethinking-attack-success-rate-comparisons-in-ai-red-teaming
    > rate (ASR) comparisons. We show, through conceptual, theoretical, and empirical contributions, that many conclusions are founded on apples-to-oranges comparisons or low-validity measurements. Our arguments are grounded in asking a simple question: (When) can attack success rates be meaningfully compared? To answer this question, we draw on ideas from social science measurement theory and inferential statistics, which, taken together, provide a conceptual grounding for understanding when numerical values obtained by quantification of system attributes can be meaningfully compared. Through this lens, we articulate sufficient conditions under which ASRs can be meaningfully compared. Using jailbreaking as a running example, we provide examples and extensive discussion of apples-to-oranges ASR

## Timeline

- 2026-08-27T21:45:50.000Z: Azure OpenAI Service delivers enterprise-ready generative AI featuring powerful models from OpenAI, enabling organizations to innovate with text, audio, and vis Companies choose Azure OpenAI Service for built-in data privacy, regional/area/global flexibility, and seamless integration into the Azure ecosystem including F Microsoft Research argues that many attack success rate (ASR) comparison conclusions in AI red teaming are founded on apples-to-oranges comparisons or low-valid