Signal Desk · Policy

Microsoft says Perception uses AI agents to test cyber defenses

2026-11-18 · 1 source · Business Standard Budget

In short

Microsoft said Perception uses specialised AI agents to simulate attacks, look into threats and help fix security holes.

What the reporting says

Microsoft said Perception uses specialised AI agents to simulate attacks, investigate threats and help fix vulnerabilities, with human oversight and controls over automated security actions. The move places the product alongside broader AI cybersecurity efforts by OpenAI, Anthropic and others. The reported focus is on using AI to strengthen security workflows while keeping controls in place.

Why it matters

Microsoft said Perception uses specialised AI agents to simulate attacks, investigate threats and help fix vulnerabilities, with human oversight and controls over automated security actions. The announcement places Microsoft within a wider AI cybersecurity push involving OpenAI, Anthropic and others. The reported relevance is to security teams looking for automated testing and threat investigation tools with guardrails.

Touches: Microsoft, OpenAI, Anthropic, AI, Companies, Global

Sources