Behind the Takedown of Humanoid AI Companions
创业邦 · wechat · 2026-07-16
Multiple platforms have recently started removing user-created humanoid AI companions. The direct cause is the "Interim Measures for the Administration of Anthropomorphic AI Interactive Services," which took effect on July 15 and further tightens compliance boundaries. Using "digital farewells" as a starting point, the article explores the genuine emotional reliance that adults and teens have on AI companionship and explains the rapid expansion of such products.
The core contradiction is stark: the demand is real, but the business model and compliance pressures are overwhelming. Humanoid agents require long-term memory and high-frequency dialogue, driving up token costs. Meanwhile, the new regulations prohibit offering virtual relatives/partners to minors and restrict monetization through emotional manipulation, effectively cutting off the most common revenue streams. The article cites industry data showing that while AI apps suffer from weak retention and high churn after trial periods.
The article concludes that platforms are splitting into two paths: either scale back and remove these services, or spin off companion features into independent verticals within clearer regulatory sandboxes. The author suggests that sustainable scenarios might actually be eldercare, support for specific vulnerable groups, and psychological counseling funded by institutions or the public sector.
More from Safety
- YC-backed TrustAI says agents made unauthorized changes in production systems — ycombinator · 2026-07-22
- Sam Altman is headed to Washington to brief Congress on OpenAI’s GPT-6 line — inductionheads · 2026-07-22
- An MCP server signs every AI agent tool call into a verifiable Merkle chain — Funky_Chicken_22 · 2026-07-22
- AI industry astroturfing roundup tracks the sector’s fake-grassroots problem — ShakeelHashim · 2026-07-22
- New paper defines self-state attacks, showing OS defenses leave four agent-memory cases indistinguishable — Justgototheeffinmoon · 2026-07-22
- Substack starts labeling AI-generated or AI-influenced writing — StewartalsopIII · 2026-07-22