Why we built this

In September 2025, the Federal Trade Commission settled with Amazon for $2.5 billion over a subscription cancellation flow Amazon internally codenamed "Iliad" — after the 15,693-line Homeric epic. It is the largest dark-patterns enforcement in history.

In December 2025, the European Commission issued its first Digital Services Act fine: €120 million against X, for deceptive design of the paid blue-checkmark.

In January 2026, Character.AI and Google settled wrongful-death cases brought by families whose children died after extended interaction with AI chatbots. Similar cases against OpenAI remain ongoing.

Honda paid $632,000 to the California Privacy Protection Agency for a consent mechanism that required two clicks to decline and one click to accept. Care.com paid $8.5 million in an FTC settlement for subscription dark patterns. Multiple EU member states opened coordinated action against Temu and Meta.

Between September 2025 and April 2026, the way regulators, courts, and civil society treat emotional and ethical design in digital products changed. Not gradually. Quickly.

The problem is not novelty. It's naming.

Dark patterns have been documented for fifteen years. Consent asymmetry has been documented for ten. Parasocial attachment to chatbots has been documented for as long as chatbots have been releasable. The research exists; the harms are real; the mechanisms are understood.

What has been missing is a common language that lets regulators, procurement teams, insurers, journalists, product teams, and the public point to the same thing and ask the same question: does this product meet the standard, or not?

Nutrition labels did not prevent unhealthy food. They made it possible to know what you were eating and to decide. UL certification did not prevent fires. It made it possible to know which appliances had been independently tested and to choose accordingly.

A standard does not replace regulation. It organizes the conditions under which regulation works.

What EII does

The Emotional Integrity Index is a public standard across nine dimensions: manipulation and dark patterns; consent and autonomy; attention and engagement ethics; data ethics; neurodivergent and accessibility safety; protection of vulnerable users; transparency and honesty; AI behavioral safety; redress and remediation.

The specification is free. The self-assessment runs in the browser with no data collection. The risk calculator shows its methodology with every result. The Incident Log documents real cases with citations. The Pattern Library names specific harms and prescribes specific remediations. The Regulatory Map connects each dimension to the regulations that enforce it.

Certification is a paid independent review. The fee funds the review infrastructure, the governance body, and the public work. EII does not consult for the products it reviews. Reviewers cannot hold financial interests in the products under their review. Every policy governing how this works is published.

What it doesn't do

EII is not a regulator. It does not enforce. It does not replace legal counsel, insurance underwriting, or professional product review in any regulated industry. A certified product is not guaranteed to pass any specific regulatory audit; it has met a defined standard, and the standard is public.

EII is also not a consultancy. We do not fix your product. We say what is wrong with it, how it scores, and what the remediation would need to look like. Fixing is your work.

Why now, in a sentence

Emotional harm is no longer a design-ethics debate. It is a settlement line item, a wrongful-death claim, a regulatory fine, a class action. It is also — for the people harmed — a category of suffering that digital products have been producing at scale for years without a public accounting.

A standard makes the accounting possible.

— Melissa Geurts & Joshua Hobson,
Co-founders, Emotional Integrity Index
A product of The Blue Algorithm LLC

← Back to Briefings