The UK government has circulated a draft guide titled "Using AI ethically and sustainably" for its employees. Among its practical tips is a reminder that users do not need to thank chatbots, a small but symbolic step toward reducing the computational overhead of every interaction. The guide also recommends using lighter model variants—such as Gemini Flash instead of Gemini Pro, or GPT Instant instead of GPT Thinking—and keeping prompts as short and direct as possible.
Beyond efficiency, the draft stresses accountability. Staff are told they are responsible for any content or decisions made with AI assistance, and they must verify outputs for factual accuracy, fairness, and the absence of discriminatory language. The document also notes that individual departments may have their own AI policies, and employees should follow those where they apply.
The guidance is still in draft form and not yet final, so specifics could change. The source does not mention any other agencies or parallel efforts, so there is no external comparison to make. The main takeaway is that even a short set of rules can nudge large organisations toward more mindful AI use—without the pleasantries.