Adult Film Producer Unmasks Prolific 'John DOE' Torrent Pirate as Meta Executive
An adult film studio has identified a prolific BitTorrent pirate as a Meta/Reality Labs executive, alleging he downloaded nearly 20,000 files — including VR porn — from both corporate and home IPs, possibly for AI training or device compatibility testing. Commenters question how someone in such a position could take this risk without basic safeguards like VPNs, and debate whether this behavior reflects individual recklessness, systemic lax security, or Meta’s broader ethics around data and copyright. The case also reignites arguments over aggressive copyright enforcement by firms like Strike 3, the legal status of using pirated material for AI training, and how liability should be assigned for activity tied to residential versus corporate networks.
Identity and behavior of the downloader
- Many are surprised someone senior at Meta/Reality Labs would torrent from a home or corporate IP instead of using better OPSEC (e.g., VPN, seedbox).
- Some suggest a family member could be responsible, but others note the timing of activity across Meta’s corporate network and the residential IP makes that less plausible, though causality is unclear.
- Others posit the person might simply be technically careless or feel effectively immune due to their status.
Corporate vs personal liability
- Key debate: was this personal piracy, or work-related activity for Meta (e.g., VR porn compatibility, moderation training)?
- Several note executives rarely do hands-on data acquisition; if this really was “for work,” it’s odd the exec did it directly, which raises questions of plausible deniability.
- Commenters stress that even if later activity shifted to a residential IP, Meta is still on the hook for earlier corporate-IP infringement.
AI training and VR/testing explanations
- Thread discusses two main “work-related” theories:
- Testing VR headsets with adult content to ensure compatibility.
- Gathering data to train models for porn detection or other AI use.
- Skeptics argue Meta could have licensed content or hired third parties instead of mass torrenting.
Scale of downloads and plausibility of personal use
- The numbers (nearly 20,000 files, thousands of porn/VR titles, ~150 downloads per day, including big “mega packs”) lead many to doubt this is normal consumption.
- Some counter that digital hoarding is common; the goal can be collecting, not watching.
Copyright, enforcement, and Strike 3
- Strike 3 is characterized by some as a large-scale copyright-troll operation; others say rights holders are entitled to enforce.
- Debate over legality of using pirated material for AI training: copying to obtain training data is still infringement even if model outputs might be lawful.
- Seeding via BitTorrent is flagged as especially risky because it clearly involves redistribution.
OPSEC, VPNs, and attribution
- Many are baffled that anyone in North America would torrent without a VPN.
- Discussion contrasts:
- Corporate IPs, where organizations can deflect blame to “unknown users.”
- Residential IPs, which are easier to tie to a household, though not to a specific person.
- Some note VPNs can reduce deniability if accounts are tied to identities; others rely on “no logs” multi-user VPNs as protection.
Meta culture, hiring, and ethics
- Several comments tie this case to prior reports that Meta torrented large datasets (e.g., books) to train LLMs, seeing a pattern of “corporate lawlessness.”
- There is broader criticism of Meta’s ethics, but also a side debate on separating bad actors from potentially beneficial acts (e.g., open-sourcing models).
Moral questions around porn datasets
- Some see little moral difference between training on porn and other media; others note added risks:
- Re-identification of performers.
- Training on coerced or illegal content (e.g., references to GirlsDoPorn).
- There is also mention of difficult, but necessary, work in training systems to detect CSAM, which some view as morally vital despite its disturbing nature.