Insights

What George Orwell’s Six Rules Can Teach Us About Context Engineering

Tom Reynolds
July 3, 2025

Context engineering is one of the most important (and often misunderstood) concepts in AI today. While prompt engineering is about how you phrase a task, context engineering is about everything that surrounds it: selecting what information to include, defining the tone, and structuring the inputs so the AI can perform the task effectively. 

At Factor, we’ve seen how powerful this shift is. Our proprietary AI tools don’t just respond to instructions: they reason across entire documents, extract relevant clauses, synthesize across policies and playbooks, and deliver precise results. That level of output doesn’t happen by accident. It happens because the context has been carefully designed to guide the AI, not just the prompt. 

That’s why we created the Legal AI Prompt Lab. It gives legal professionals practical tools to get better results with AI, starting with downloadable prompt cards and walkthrough videos that provide real-life, immediately applicable examples. The Prompt Lab Library is a free subscription that releases a new case each month, covering tasks like analyzing unfamiliar contracts, reviewing marketing content for legal risk, and turning messy notes into clear legal memos. It's already helping legal teams work faster, clearer, and with more confidence, and shows how best practices can be adapted to fit their own workflows. 

When we think about how to interact with AI effectively, George Orwell’s famous six rules for writing remain strikingly relevant. Though they were written long before AI existed, they offer a simple and powerful guide to better context engineering. 

1. Never use a long word where a short one will do. 

Clear, simple instructions reduce ambiguity. In AI workflows, precision always beats flourish. 

2. If it is possible to cut a word out, always cut it out. 

Context windows are limited. Every input should earn its place. Strip away filler and keep what matters. 

3. Never use a metaphor, simile or other figure of speech you are used to seeing in print. 

AI interprets inputs literally. Vague or figurative language leads to vague or generic outputs. 

4. Never use the passive where you can use the active. 

Direct commands like “Extract key risks” or “Compare termination clauses” guide models more effectively than abstract or passive phrasing. 

5. Never use a foreign phrase, a scientific word or a jargon word if you can think of an everyday English equivalent.

 Avoid legalese unless it’s essential. Plain language improves model understanding and output consistency. 

6. Break any of these rules sooner than say anything outright barbarous. 

Context engineering is part art, part science. Be guided by clarity, not rules for their own sake. 

These principles are embedded in the way we build and refine AI at Factor. Our tools are shaped by the needs of real users and tested by over 400 legal professionals — which means they reflect how legal work actually happens, not just how it's theorized. Prompt engineering is a starting point. Context engineering is what delivers reliability, consistency, and outcomes that hold up in the real world. 

The Prompt Lab is your starting point for learning how to interact with AI this way. You’ll find focused guides, realistic examples, and walkthroughs that show how to set up your inputs for success. And soon, we’ll be adding resources to help you go a step further — with guidance on how to build your own legal AI agents. 

If you're wondering: yes, AI can even help you write a better AI prompt. Just ask it to apply Orwell’s six rules to your draft. Clearer, faster, better… and maybe just a little Orwellian. 

Access the Prompt Lab today and start putting context engineering into practice.