Proton Spam and the AI Consent Problem
Key Points
- 1The author details receiving unsolicited "Lumo AI" emails from Proton despite explicitly opting out, which Proton initially miscategorized as a "business newsletter" instead of a Lumo update.
- 2This incident is framed within a broader critique of the AI industry's disregard for user consent, citing similar issues with GitHub Copilot emails and general non-acceptance of "no."
- 3Following escalation, Proton's Head of Customer Support and CTO admitted the email was sent due to a bug in their system and apologized, reaffirming their commitment to communication consent.
The provided text details a user's experience with unsolicited marketing emails, specifically highlighting a perceived disregard for user consent by technology companies, particularly concerning Artificial Intelligence (AI) related products.
The core problem originates with Proton, a privacy-focused company, sending an email promoting its "Lumo" (AI) offering to the author. This occurred despite the author having explicitly unchecked the "Lumo product updates" toggle in their email subscription settings, indicating a clear desire to not receive such communications. The author, a paying business customer, considers this email "spam" and a violation of GDPR and UK data protection laws, as well as an abuse of Proton’s own service.
Upon contacting Proton support, the initial response directed the author to the very opt-out setting they had already configured. Subsequent exchanges involved support requesting screenshots and eventually providing a justification: the email was allegedly part of a "Proton for Business newsletter" promoting "Lumo Business Suit" for business plans, distinct from the "Lumo product updates" category. Proton support claimed "Product Updates and Email Subscription are two different things," which the author dismissed as "baloney" given the email's content and sender address ("From Lumo," @lumo.proton.me).
The author broadens this incident into a general critique of the AI industry's "non-consent" problem. They argue that the AI industry consistently disregards user preferences, copyright, and even engages in aggressive online behaviors (e.g., AI bots DDoS-ing websites). Examples beyond Proton are cited, including Mozilla/Firefox and a subsequent incident with GitHub (owned by Microsoft), where the author received an email promoting "GitHub Copilot SDK" despite having never opted into GitHub newsletters and having disabled all possible email notifications. The "Unsubscribe" link on GitHub revealed a hidden newsletter list where "GitHub Copilot emails" were unchecked but still sent. This reinforces the author's central claim that tech companies, especially those involved with AI, actively opt users in or bypass explicit opt-out choices.
Following the publication of the author's blog post, Proton provided further clarification and a resolution. A "Specialist Support / Mail Delivery (Engineering) Team" initially explained that "Overlapping Categories of notifications" meant Lumo communications could fall under both "Product Updates" and "Email Subscriptions (Announcements, Newsletters, and Promos)," leading to the email despite the specific opt-out. However, after escalation to Proton's Head of Customer Support, the company formally apologized, acknowledging that the author should not have received the newsletter. The Head of Customer Support confirmed a "bug in our system" that their technical team was resolving, reiterating Proton's commitment to "communication consent." This "bug" explanation was corroborated by Proton's CTO on Hacker News. While accepting the explanation, the author still expresses concern over the initial handling and the broader trend of consent disregard in the tech industry.