This week we're reading an unusually thoughtful op-ed on the question of overarching AI regulation, data privacy concerns related to telehealth platforms and reflecting on how in-house ethics work operates within Big Tech – and what we can realistically expect from it.
There’s Only One Way to Control AI: Nationalization | Politico
One of the better op-eds we've read on AI. Although the policy suggestion (an AI governing body) is not inherently novel, the thinking behind it expertly merges technologist and policy expertise. Well worth a read.
The Hidden Cost of Teletherapy | Tech Policy Press
It's hard to imagine any personal data being more sensitive than mental health data. This report on the marketing of data from teletherapy platforms – which are not covered by HIPAA – is deeply alarming. The piece lies out the stakes of data privacy in compelling terms (even with HIPAA protected data, "the idea of consent holds virtually no meaning given the lack of consumer awareness and choice") and offers some excellent recommendations.
Walking the Walk of AI Ethics: Organizational Challenges and the Individualization of Risk among Ethics Entrepreneurs
With every AI startup touting their commitment to "responsible of AI" and Meta pointing to recent studies as evidence that "there is little evidence that key features of Meta’s platforms alone … have meaningful effects on key political attitudes, beliefs or behaviors," this paper raises some excellent questions as to what leverage in-house ethics teams have at technology companies.
Pro-Wagner Accounts Keep Showing Up On Facebook and Instagram after Meta's Ban On the Mercenary Group, Report Shows | CNN Business
In this week of social media whack-a-mole, researchers at the Institute for Strategic Dialogue found that pro-Wagner group content not only persists on Meta-owned platforms in contradiction of their policy against "dangerous organization," but appears to be promoted by the Meta algorithms.