AI Policy

Summary

Ethical Consumer, we believe technology should serve people and planet, not exploit them. The
rapid growth of artificial intelligence (AI) raises urgent ethical concerns: from mass data
extraction and surveillance, to labour rights violations, environmental damage, and the
entrenchment of corporate power.

Our stance:

• We reject AI systems that exploit workers, communities, or ecosystems.
• We will not blindly embrace AI – our approach will be critical and we’ll only use it where
  we are sure of its benefits.
• We oppose technologies that deepen inequalities or erode accountability.
• We continue to prioritise human-led research and writing.
• We reject the prioritisation of frictionless efficiency, and appreciate the value of slow
  learning and human analysis.

We use AI tools cautiously and critically, always guided by our values. Where AI undermines our
mission, we will resist.

Ethical Consumer exists to imagine and build a fairer economy, including aligning with degrowth
principles. AI, like any tool, must be judged on its practice, not on its promise.

Please note this AI policy is subject to change as new research emerges, technology evolves
and our understanding deepens. We reserve the right to update this page at any time.

Purpose

This policy sets out Ethical Consumer’s position and approach to AI across our publishing,
campaigning, research, and organisational work.

Definition

For the purposes of this policy, “AI” refers to machine learning and generative systems (e.g. text,
image, video tools) that automate or augment human tasks.

Risks identified

• Labour exploitation: hidden workforces are often paid below living wages.
• Environmental costs: massive energy and water use of large-scale data centres powering
  AI models.
• Bias and inequality: reinforcement of systemic discrimination in datasets and outputs.
• Corporate capture: concentration of power in a handful of tech giants, undermining
  democratic control.
• Misinformation: potential to suppress truth and to erode trust and truth in public discourse.
• Deskilling: outsourcing writing and critical thinking skills.

Opportunities identified:

• Faster Research: AI could help us process external corporate reports more efficiently.
• Database Enhancement: AI could analyse trends offering new insights to readers.
• Process mining: AI could help us refine internal resource allocation and identify bottlenecks.
• Admin: AI could help with automations, translations and supporting accessibility.

Ethical Consumer stance

• We will not adopt AI uncritically and we will resist AI applications that conflict with our
  mission of social and environmental justice.
• We may experiment with limited, transparent use of AI tools where they support our work
  without undermining our values.
• Where we do use AI it will always have human oversight.
• We will prioritise open-source, ethical, and worker-friendly tools wherever possible.
• We reject AI systems that conflict with our mission of creating a more ethical economy. This
  may mean campaigning against specific models, ownerships or approaches and for robust
  regulation too.
• We will only use AI as a support tool not as a substitute for human authorship.
• We will always disclose if AI is used in producing finished content.
• We will fact-check all AI outputs against reliable human sources.
• We will consider environmental impact before extensive use.

Accountability & review

• This policy will be reviewed annually to respond to emerging risks and debates.
• Coop members and board members can raise concerns about AI use at any time.
• We will publish updates on our approach to AI to remain accountable to our members and
  community.

Permitted

• Spellcheck, grammar, and readability checks.
• SEO assistance.
• Admin assistance.
• Data merging, translations and background research checks.
• Speech-to-text/text-to-speech transcriptions for accessibility requirements.

Limited/Conditional

• Generating limited marketing copy drafts.
• Assisting in data analysis of large datasets.
• Background ideation for creative projects (never final content).
• AI tools integrated into third party software that ECRA uses except where it contradicts
  ECRA’s overarching AI stance.

Not Permitted

• Writing finished articles or editorial content.
• Generating AI images, video, or audio for publication.
• Resharing AI images, video, or audio for publication (where they are labelled as being AI).
• Deciding scores, final ratings, or “Best Buy” recommendations.
• Any use that risks replacing human editorial judgement.