Our position on automation and AI
Clear rules where they are sufficient, selective AI use and care for the resources involved.
At elOS, we favour clear rules for tasks that can be defined reliably, and selective use of AI where it provides a useful benefit. Reliability, the time people spend using a tool and the resources needed to run it belong in the same decision.
Start with the task
A calculation, a required field or an access permission can usually be handled by explicit instructions. This is often called deterministic programming: the same inputs and relevant conditions produce the same result. It gives a team a defined behaviour to test and maintain.
AI can be useful when a task involves language or variation that is difficult to describe through fixed rules. Its value depends on the quality of the result, the consequences of an error and the effort needed to check it. Adding AI should improve the task enough to justify those demands.
The UK government's Data and AI Ethics Framework recommends exploring alternatives, including approaches without AI, before choosing a technical solution. That is consistent with our position: choose the method that serves the need.
Resources are part of the choice
Digital services rely on physical infrastructure. The International Energy Agency estimates that data centres used 415 TWh of electricity in 2024, around 1.5% of global electricity consumption. That figure covers data centres as a whole, including services beyond AI. Source: IEA, Energy and AI, 2025.
This context strengthens our case for avoiding unnecessary processing. It does not establish the environmental cost of an individual task, which depends on the technology and how it is used.
Rights and responsibility
The EU’s AI Act sets obligations according to the risks of different AI uses, including transparency and human oversight where required.
In Brazil, the Senate-approved version of PL 2338/2023 proposes rights to information about interacting with AI, privacy and protection against unlawful discrimination. For people affected by high-risk systems, it also provides for explanations, challenges to decisions and human review, subject to the context and associated risks. It requires impact assessments for high-risk systems and includes provisions to reduce energy and other resource use in general-purpose and generative AI. It remains a bill under consideration, rather than an enacted law. Source: the Senate’s explanation of the approved text.
These principles inform our position: choose AI for a defined purpose, consider its effects on people and use resources responsibly.
The future we want to support
We see value in technology that helps communities care for their surroundings, supports viable local organizations and gives people more room to cooperate. Progress should be visible in the work people can sustain and the resources they can use more thoughtfully.
For elOS, that means favouring understandable tools, avoiding unnecessary processing and building around the people doing the work. AI has a place when it serves those aims. Our ambition is for useful technology to become accessible to teams whose contribution is much greater than their budget.