As AI Coding Gains Popularity, Microsoft Develops AI to Counter Cyber Threats
Developments in artificial intelligence (AI) have accelerated software development. Unfortunately, they have also increased the risk of security vulnerabilities emerging, as more developers rely on AI coding assistants to write and repair code.
Cybersecurity experts warn that without adequate protection and verification processes, the use of AI can magnify the potential for software vulnerabilities.
In response to this situation, Microsoft is developing a specialised AI model to detect and fix security vulnerabilities. The move is intended to counter cyber threats that are growing and increasingly exploiting AI technology.
AI Accelerates Development, but Risks Rise Alongside It
The use of AI in software development is becoming more widespread, as AI can help generate code, find bugs, and automate various programming tasks.
However, AI is also being used by cyber criminals to find security gaps more quickly.
In response, Microsoft introduced the MAI-Cyber-1-Flash cybersecurity model within its Project Perception platform, designed to help analysts detect, prioritise, and repair software vulnerabilities more efficiently, allowing organisations to be better prepared for continuously evolving cyber threats.
Behind its benefits, AI as a programming assistant also carries security risks.
Research by cybersecurity firm Pillar found that a number of popular AI coding agents, such as Cursor, Codex CLI, Gemini CLI, and Antigravity, are vulnerable to sandbox escape attacks, which allow attackers to execute malicious commands outside a safe execution environment through the manipulation of project files.
Although most developers have released updates to close these gaps, Pillar considers conventional security approaches inadequate for facing AI agents capable of influencing systems through a series of automated actions.
The growing use of AI in programming means that a single security flaw could potentially spread to many applications if the code generated is not secure.
For this reason, experts stress that AI must be treated as an aid, whilst every piece of code must still undergo audits, security testing, and vulnerability scanning before use.
Amid these conditions, the cybersecurity industry is entering a new chapter, as technology companies harness AI to strengthen their defences whilst cyber criminals use the same technology to find and exploit software weaknesses.