2 ms·
Applying laws based on MAUs doesn't seem like a reasonable approach. MAUs are a self-reported and subjectively defined metric and the MAU threshold is arbitrari
by root_axis 6y ago
Applying laws based on MAUs doesn't seem like a reasonable approach. MAUs are a self-reported and subjectively defined metric and the MAU threshold is arbitrarily chosen. Technology laws should not be applied based on apparent popularity.
I think a better version of this legislation would apply to any corporation that collects PII or collects browser activity outside of 1st party domains, in order to serve ads to users.
- munfred 6y agoThis has actually also be proposed by Mark Warner in 2018 on a whitepaper titled “Potential Policy Proposals for Regulation of Social Media and Technology Firms”: https://www.warner.senate.gov/public/_cache/files/d/3/d32c2f17-cc76-4e11-8aa9-897eb3c90d16/65A7C5D983F899DAAE5AA21F57BAD944.social-media-regulation-proposals.pdf https://www.warner.senate.gov/public/_cache/files/d/3/d32c2f... It discusses 20 topics for potential legislation, listed below. Look at items 13 and 16 in the whitepaper, they go very much along the lines of what you said. 1) Duty to clearly and conspicuously label bots 2) Duty to determine the origin of posts and/or accounts 3) Duty to identify inauthentic accounts 4) Make platforms liable for state-law torts (defamation, false light, public disclosure of private facts) for failure to take down deep fake or other manipulated audio/video content 5) Public Interest Data Access Bill 6) Require Interagency Task Force for Countering Asymmetric Threats to Democratic Institutions 7)Disclosure Requirements for Online Political Advertisements 8) Public Initiative for Media Literacy 9) Increasing Deterrence Against Foreign Manipulation 10) Information fiduciary 11) Privacy rulemaking authority at FTC 12) Comprehensive (GDPR-like) data protection legislation 13) 1st Party Consent for Data Collection 14) Statutory determination that so-called ‘dark patterns’ are unfair and deceptive trade practices 15) Algorithmic auditability/fairness 16) Data Transparency Bill 17) Data Portability Bill 18) Interoperability 19) Opening federal datasets to university researchers and qualified small businesses/startups 20) Essential Facilities Determinations