Artifact 3 Live AI App Available

AI Shield

A friendly AI-powered cybersecurity assistant for small business owners — clear, calm, and actionable guidance without jargon.

AI Shield — AI-powered cybersecurity assistant
Now Live

🛡️ Try AI Shield in Action

The concept is now a fully functional AI assistant powered by Claude. Ask it about phishing, ransomware, passwords, 2FA, data breaches — in plain English. Real answers, real protection. Bring your own Anthropic API key or explore in demo mode.

Claude API Live Chat Threat Dashboard Plain English Guidance
Launch AI Shield

No signup · Key stays in browser

Overview

AI Shield is built for non-technical users. Think of it as a friendly security coach that explains risks in plain English and guides quick, confident fixes — without shaming anyone for not knowing the jargon.

Introduction
AI Shield is an AI-powered cybersecurity assistant built specifically for small business owners and non-technical users. It monitors your digital environment, explains potential risks in plain English, and guides you through resolving them — no jargon required. The live version uses Claude to provide real, context-aware security guidance.
Description
A prioritized threat dashboard surfaces what matters: risky permissions, suspicious behavior, weak passwords, and simple guided remediations. A live AI chat — powered by Claude — answers any security question in conversational, plain-English terms. The built-in knowledge base (accessible via quick-topic buttons) covers 2FA, malware, ransomware, phishing, backups, cloud security, and more.
Objective
Empower small businesses with smart, easy-to-understand tools that reduce risk without requiring technical expertise. The goal is to make cybersecurity feel approachable — not intimidating — so that real protective action actually gets taken.
Process
User research with 25+ small business owners → low-fi prototypes focused on clarity and empathy → usability testing → high-fi mockups in Figma → front-end prototype with placeholder AI → scenario testing with GPT prompts → live Claude API integration with a custom security-focused system prompt. Key insight: security is emotional — people want to feel safe, not shamed.
Tools and Technologies Used
  • HTML, CSS & JavaScript – Clean, responsive UI for desktop and mobile.
  • Anthropic Claude API – Powers the live AI assistant with a custom cybersecurity system prompt.
  • GitHub Pages – Live prototype hosting — zero build step, fully static.
  • Figma – Wireframes & high-fi designs with accessibility focus.
  • JetBrains Mono / Syne – Typography chosen for a tactical, technical aesthetic.
  • Demo Mode – Pre-written curated responses for the most common scenarios (no key needed).
Value Proposition
Expert-level protection in an approachable interface. It focuses on three questions: What's the risk? Why does it matter? What should I do now? Even a solo founder can protect their business in minutes a week. The dynamic threat score and color-coded alert panel help prioritize what to address first.
Unique Value
A supportive tone and coaching mindset. Real-world analogies, adaptive language, and bite-sized guidance with each alert help users learn while staying protected. The live AI understands context — ask a follow-up and it remembers what you said earlier in the conversation.
Relevance
With remote work and SaaS sprawl, small businesses face growing threats without enterprise budgets. 43% of cyberattacks target small businesses, yet most lack even basic security practices. AI Shield bridges that gap — bringing the knowledge of a security expert into a friendly, affordable, always-available assistant.
References
← Back to Projects Launch AI Shield