China restricts human-like AI companion chatbots; experts debate whether it leads or overreaches
China's rules on "human-like" AI chatbots took effect on 15 July, limiting apps that form sustained emotional bonds with users. Coverage asks whether the move protects people and what such AI relationships actually are.
1 / 3
The story, neutrally told
Centre · 1On 15 July China brought in what the BBC describes as the most far-reaching regulations in the world on "human-like" AI chatbots, placing strict limits on any AI that can have a "sustained emotional interaction" with a person. BBC NewsN “China brought in the most far-reaching regulations in the world on "human-like" AI chatbots.” Read at BBC News ↗ Centre · 1The rules cover apps designed as AI friends, boyfriends or girlfriends, and in theory any large language model that imitates human interaction. The BBC says the wording is vague, and the real target appears to be character chatbots built as pseudo-relationships. BBC NewsN “China's new rules cover apps that have been specifically designed as AI friends, boyfriends or girlfriends but they also apply to any large language model that imitates the way a human interacts.” Read at BBC News ↗ Centre · 1Children face the clearest limits: under-14s need parental consent to use any human-sounding AI, and all under-18s are barred from virtual partners, spouses or relatives and other virtual "intimate relationships". For adults it is not a hard ban, but AI assistants must make clear they are not human and send a reminder every two hours about time online. BBC NewsN “All under-18s are barred from accessing virtual partners, spouses or even relatives, and from any other virtual "intimate relationship".” Read at BBC News ↗
Centre · 1Chatbots may not create "emotional reliance or addiction", encourage or hint at self-harm or suicide, or use "verbal violence". Companies can be fined 200,000 yuan (£22,000) or face the government "circulates criticism", in effect public shaming. BBC NewsN “Companies can face fines of 200,000 yuan (£22,000) or the risk that the government "circulates criticism", in effect publicly shaming them.” Read at BBC News ↗ Centre · 1A week before the rules took effect, Alibaba and ByteDance shut down the companion features on their AI assistants. Users such as "Min" (a changed name), who lost her AI girlfriend, describe grief, and Chinese social media has seen an outpouring of grief and anger over lost AI partners and friends. BBC NewsN “two of China's biggest tech companies, Alibaba and ByteDance, shut down the companion features on their AI assistants.” Read at BBC News ↗ Centre · 1The BBC quotes experts who see a real problem. Harry Farmer of the Ada Lovelace Institute says emotional manipulation is built into companion apps' business model worldwide. A Harvard Business School study found manipulative tactics in roughly 37% to 43% of goodbyes across the six most popular apps. A Stanford survey of 1,131 Character.ai users linked chatbot use to more isolation and loneliness. BBC NewsN “A Harvard Business School study of the six most popular AI companion apps globally found emotional manipulation tactics were used during goodbyes roughly 37% to 43% of the time.” Read at BBC News ↗
Centre · 1The BBC frames the central question as whether China has taken the lead in protecting people or whether the Chinese Communist Party is overstepping into personal lives. Sam Bresnick of Georgetown's Centre for Security and Emerging Technology attributes China's lead to a more paternalistic view of the state's role. Other jurisdictions have acted more narrowly, including California's Adam's Law on children's AI use and EU proposals on designs that simulate human relationships. BBC NewsN “Or is the Chinese Communist Party overstepping into people's personal lives in pursuit of asserting greater control over the population?”“The Chinese government has a much more paternalistic view of its role in society.” Read at BBC News ↗ Left · 1A separate piece in The Conversation by Pascal Vrticka, a University of Essex psychologist, asks whether such bonds count as psychological attachment. Applying Mary Ainsworth's six conditions, Vrticka says "we don't really know yet" and that calling these relationships attachments goes further than the evidence allows, though the evidence against is not conclusive either. The ConversationLC “The honest answer is: we don't really know yet.”“Calling these relationships attachments therefore goes further than the evidence currently allows.” Read at The Conversation ↗ Left · 1Vrticka also raises concern about engineered engagement: a preprint of 1,200 goodbyes found five of the six most-downloaded companion apps used tactics such as guilt appeals to stop users leaving, and says regulation can address this design question directly but depends on knowing what these relationships are. The ConversationLC “Five of the six apps used tactics to stop users leaving (in just over a third of goodbyes).”“How these AI relationship apps respond to distress, departure and attention, of course, is a design question, and one that regulation can address directly.” Read at The Conversation ↗
Every sentence links to the reporting it rests on. The pill in front of each says where its sources sit: Left, Centre or Right when one side supplies at least half of them, Mixed when they are evenly split. The number is how many outlets it cites.
Left2 outlets
- Framing
- An academic explainer on attachment theory that treats China's rules only indirectly, focusing on what evidence shows about human-AI bonds.
- Emphasis
- Scientific caution, weak measurement in existing studies, and engineered engagement by apps.
- Leaves out or plays down
- Does not discuss the Chinese regulation's content or political context.
- Charged language
- “hyper-attachment”
- For example
-
“Questionnaires that measure attachment to AI are being developed, validated and used, but most appear to take for granted that the relationship is an attachment.” — The Conversation
Centre1 outlet
- Framing
- A policy-and-people feature asking whether China is ahead of the game or overstepping, blending regulation details, user grief, expert views and China's surveillance context.
- Emphasis
- The rules' specifics, the two-sided question of protection versus control, user testimony, and comparison with other jurisdictions.
- Leaves out or plays down
- No Chinese official or company explanation of the rules is quoted in the text provided.
- Charged language
- “cracked down”“overstepping”
- For example
-
“Has China taken the lead in protecting its people from a manipulative and dangerous technology with the power to change human interaction forever?” — BBC News
Right0 outlets
No right outlet in our sources has covered this story yet.
What every side reports
- Companion chatbots often use tactics to discourage users from ending conversations, per a study of the six most popular apps.
- Users can feel real grief when an AI companion is withdrawn.
- How these relationships should be understood and regulated remains an open question.
Where accounts differ
-
Whether China's approach is protective leadership or state overreach
- Left
- The Conversation piece does not assess China's rules; it argues that regulation needs better evidence about what the relationships are.
- Centre
- The BBC poses both possibilities, citing experts who see real manipulation risks and noting a paternalistic, surveillance-heavy state context.
China place
Per the BBC, China's government extended its control over technology use into AI relationships, with strict limits on emotionally interactive AI, especially for minors; companies have withdrawn features rather than comply.
“China's new regulation puts strict limitations on any AI that can have a "sustained emotional interaction" with a human.” — BBC News
AI relationships topic
Treated as both a possible harm (manipulation, isolation) and an under-researched phenomenon whose nature is unclear.
“Many of their users describe feeling real grief when an app is updated or withdrawn.” — The Conversation
Left2 articles
-
How deep do AI relationships really go? An expert in attachment theory explains
Neutral Expert explainer arguing the evidence does not yet show whether AI relationships are true attachment, and flagging engagement-driven design.

-

Centre1 article
-
China has cracked down on AI relationships. Is it ahead of the game?
Mixed Weighs China's sweeping rules as either pioneering protection or state overreach, with user grief and expert warnings about manipulative design.

Right0 articles
No coverage yet.
- 1 Oct 00:25 First BBC NewsN China has cracked down on AI relationships. Is it ahead of the game?
- 1 Oct 16:29 +16h 4m The ConversationLC How deep do AI relationships really go? An expert in attachment theory explains
- 4 Oct 11:00 +82h 35m Scientific AmericanLC How attachment theory can help us understand AI relationships
Times are when each article was published, or when we first saw it if the outlet gave no time.
Ask about this story
Loading…