Microsoft Introduces New AI Models For Coding And Cybersecurity - DominzyLoaded Tech Microsoft Introduces New AI Models For Coding And Cybersecurity - DominzyLoaded Tech

Microsoft Introduces New AI Models For Coding And Cybersecurity

Microsoft Introduces New AI Models For Coding And Cybersecurity

Microsoft Introduces New AI Models For Coding And Cybersecurity
Microsoft has introduced new artificial intelligence models designed to support software development and cybersecurity. The announcement expands the company's in-house AI portfolio with specialised tools for developers and security professionals.

The new releases include MAI Code 2, a coding model designed to handle software engineering tasks, and MAI Cyber 2 Flash, which focuses on cybersecurity. Both models are built to deliver specialised performance at lower operating costs.

Microsoft announced the models at its Ignite 2026 conference on November 17. The company says they are designed to help businesses complete technical tasks more efficiently without relying entirely on expensive general-purpose AI models.

The announcement also includes MAI Code 2 Flash, a smaller coding model for routine development work. Together, the releases demonstrate Microsoft's growing focus on building AI models for specific professional tasks.

Microsoft's New AI Models for Coding and Cybersecurity

The two main releases are MAI Code 2 and MAI Cyber 2 Flash. Although both are designed for technical work, they serve different purposes.

MAI Code 2 focuses on software engineering activities such as debugging, implementing changes across multiple files and reasoning through existing code repositories. MAI Cyber 2 Flash is designed to help security professionals identify software vulnerabilities and recommend ways to address them.

Microsoft is developing these models to handle specialised workloads. This approach allows organisations to choose AI tools based on the type of work they need to complete.

What Is MAI Code 2?

MAI Code 2 is Microsoft's flagship coding model, developed to assist with everyday enterprise software engineering tasks. It is designed to handle complex coding activities, including debugging, implementing changes across multiple files and working with large software repositories.

The model can also assist with taking code towards production and developing user interfaces. Its repository-wide reasoning capabilities help it understand how different parts of a software project relate to one another.

Microsoft says MAI Code 2 achieved 98% of GPT-6.1 Sol's performance at approximately 40% of its price. The company also reported that it reached 96% of Claude Opus 5.5's performance at 5% of the cost.

These figures are based on Microsoft's evaluations across four coding benchmarks: DeepSWE 1.1, Terminal Bench 2.1, SWE-Atlas and FrontierCode v1.1. They represent company-reported results for everyday enterprise coding tasks rather than a guarantee of performance across every programming project.

Microsoft has also introduced MAI Code 2 Flash, a smaller and more cost-efficient model designed for high-volume coding activities. It can assist with explaining code, making targeted edits, fixing routine errors and supporting rapid development iterations.

Both MAI Code 2 and MAI Code 2 Flash are available through GitHub Copilot and Microsoft Foundry.

How MAI Cyber 2 Flash Supports Cybersecurity

MAI Cyber 2 Flash is Microsoft's specialised AI model for cybersecurity. It is designed to help security professionals examine software, identify vulnerabilities and understand potential threats.

Software vulnerabilities can expose applications and computer systems to attacks. Identifying these weaknesses often requires examining large amounts of code and understanding how different components interact.

MAI Cyber 2 Flash can assist security teams with:

  • Vulnerability discovery: Identifying potential weaknesses in software and code.
  • Security analysis: Examining identified issues to help security professionals understand possible risks.
  • Remediation recommendations: Suggesting ways to address vulnerabilities and improve software security.
  • Defensive security operations: Supporting security teams as they investigate and respond to potential threats.

Microsoft says MAI Cyber 2 Flash performs ahead of GPT-6 Astra on the CyberGymE2E benchmark while costing less than 2% as much per task. The company attributes its performance to its security expertise and proprietary data.

The model is part of Microsoft's broader cybersecurity strategy, which includes integrating specialised AI capabilities with its security tools. Its capabilities are intended to support defensive security work, helping organisations identify and address weaknesses before they can be exploited.

How the New Models Compare With Other AI Tools

Microsoft's announcement places its new models in competition with AI products from companies such as OpenAI and Anthropic.

MAI Code 2 targets software development tasks that developers already use coding assistants for, while MAI Cyber 2 Flash focuses on vulnerability discovery and other cybersecurity activities.

Microsoft's main selling point is the combination of specialised performance and lower operating costs. Rather than using an expensive general-purpose model for every technical task, organisations can consider models designed for particular workloads.

However, benchmark results do not establish that one model is better in every situation. Actual performance depends on the task, available tools, programming environment and complexity of the work.

Where To Access Microsoft's New AI Models

Microsoft has made MAI Code 2 and MAI Code 2 Flash available through GitHub Copilot. Developers can select the models through the available model options when using the service.

Both coding models are also available through Microsoft Foundry, Microsoft's platform for building and deploying AI applications.

MAI Cyber 2 Flash is being integrated into Microsoft's cybersecurity platform, Perception. Its availability is tied to the security platform rather than being offered as a general-purpose coding assistant.

Access may depend on the platform, subscription and availability arrangements. Developers and organisations should check Microsoft's official product information for the latest access details.

Other AI Models Announced at Microsoft Ignite 2026

Beyond coding and cybersecurity, Microsoft announced additional models for image generation, voice interaction, transcription and decision-making.

Some of the other releases include:

  • MAI Image 2.7: An image generation and editing model designed to create and modify images from instructions.
  • MAI Image 2.7 Flash: A lighter image model intended for high-volume image generation tasks.
  • MAI Realtime 1: A speech-to-speech model designed for natural, low-latency voice conversations.
  • MAI Transcribe 2: A transcription model used in Microsoft's Dragon Copilot service.
  • Microsoft-Decision-1.1: A model designed to help AI applications make decisions such as routing requests, classifying information and prioritising tasks.

These releases show that Microsoft is expanding its AI development beyond general-purpose chatbots and into specialised tools for business and professional applications.

What Microsoft's New AI Models Mean for Developers and Businesses

Microsoft's new models give developers and businesses additional options for handling software development and cybersecurity tasks. Their specialised capabilities could make it easier to select an AI tool that matches a particular workload.

Developers can use MAI Code 2 to assist with debugging, multi-file implementation and other software engineering activities. MAI Code 2 Flash provides a more cost-efficient option for routine coding tasks that require frequent interactions.

Security professionals can use MAI Cyber 2 Flash to support vulnerability discovery and security analysis. Its integration with Microsoft's cybersecurity platform also connects AI-assisted security work with the tools organisations use to protect their systems.

However, AI-generated code and security recommendations still require human review. Developers need to test generated code before deployment, while cybersecurity professionals should verify identified vulnerabilities and proposed fixes before applying them.

Latest

Kindly Join our Telegram channel & active group chats to get quicker access to the latest tech news, mobile tips, free airtime codes and free browsing cheats

Telegram Channel: Subscribe Here

WhatsApp Channel: Follow Here

Whatsapp Group: Join Here

Telegram Group: Join Here

We are now on Mi Community for those that uses Xiaomi devices. Follow us now.
Mi Community: Click Here

Follow us on Instagram
Instagram: Join Here

Subscribe to Watch our Tech Tutorial Videos on YouTube.
YouTube: Subscribe Here

Make Sure you tick the "Notify me" box, so that you get notified when the admin responds to your comments or questions.
EmoticonEmoticon