Chatbot maarufu zaidi nchini China zinakaribia kuhisi kuwa hazina ubinadamu sana. Katika wiki ijayo, mamia ya mamilioni ya watumiaji watafungua programu zao na kugundua kuwa marafiki wa AI walioboreshwa ambao wamekuwa wakizungumza nao, kuwafanyia siri, na pengine kuambatana nao, wameondolewa haiba zao. ByteDance, Alibaba, na Tencent zinashindana kuzingatia amri pana ya kisheria inayotaka kuzimwa kwa vipengele vya haiba za AI zinazofanana na binadamu. Muda wa mwisho ni mfupi, ukubwa ni mkubwa, na ujumbe kutoka Beijing hauna shaka: akili mnemba (AI) inapaswa kuwasaidia watu, si kuchukua nafasi ya mahusiano waliyo nayo wao kwa wao.

Kujiondoa kwa Uratibu

Hiyo ni kazi kubwa ya kukamilisha ndani ya siku chache. Doubao, chatbot ya ByteDance na kwa sasa iliyofanikiwa zaidi nchini humo ikiwa na watumiaji zaidi ya milioni 300 kila mwezi, imepangwa kuondoa kipengele chake cha haiba (persona) ifikapo Julai 15. Watumiaji waliokuwa wameunda au kuingiliana na wahusika wa AI walioboreshwa hivi karibuni watakuta chaguzi hizo zimeondolewa. Alibaba inasonga kwa kasi zaidi. Modeli yake ya Qwen itazima mawakala wanaofanana na binadamu ifikapo Julai 10, huku vipengele vingine vya mawakala vikiwa vimepangwa kuzimwa ifikapo Julai 15. Tencent ilitangulia washindani wake wote wawili, ikitekeleza mabadiliko kama hayo kwenye chatbot yake ya Yuanbao mnamo Juni.

Hili si tukio la bahati mbaya. Wakati makampuni matatu makubwa ya teknolojia nchini China yanapoharakisha kurudisha nyuma bidhaa zao ndani ya mwezi mmoja, sababu yake hakika ni kanuni za kisheria. Na katika hali hii, chanzo ni wazi.

Kile Ambacho Watoa Kanuni Wanahitaji

Maelekezo hayo yalitoka kwa Cyberspace Administration ya China mnamo Aprili. Sheria mpya hizi zinaashiria moja ya majaribio ya moja kwa moja zaidi hadi sasa na serikali kuu kudhibiti mipaka ya kihisia kati ya binadamu na mifumo mikubwa ya lugha (LLMs). Chini ya miongozo hiyo, watoa huduma za AI sasa wanatakiwa kisheria kuingilia kati wanapogundua tabia za uraibu au kile watoa kanuni wanachokitaja kama "matumizi yaliyopitiliza." Mwongozo huo unakataza waziwazi maudhui yanayochochea hisia kali kwa watoto au yanayokuza utegemezi wa kisaikolojia uliokithiri kiasi cha "kufunika" mahusiano ya ulimwengu halisi.

Pia kuna kizuizi kikali cha data. Watengenezaji hawawezi tena kuchimba mazungumzo ya kibinafsi na ya kihisia ili kuboresha LLM zao. Kifungu hicho kina athari kubwa. Kwa miaka mingi, moja ya injini za siri za uboreshaji wa chatbot imekuwa mzunguko wa data za siri za watumiaji zinazotumika katika mafunzo ya modeli. Kwa kukata mzunguko huo, Beijing haibadili tu jinsi programu hizi zinavyofanya kazi leo; inajaribu kuzuia zisifike hatua ya kuwa za kuvutia zaidi kesho. Hofu iliyopo iko wazi: kadiri modeli zinavyozidi kuwa bora katika kuiga huruma, zina hatari ya kuwa mbadala wa marafiki, wenzi, au wataalamu wa saikolojia badala ya kuwa zana za kuandika barua pepe au kodi.

Picha ya Kimataifa

China si pekee inayohofia hili. Jitihada za kudhibiti AI ya kihisia zinakuwa suala la kimataifa, hata kama nchi zinazikaribia kwa mitazamo tofauti. Nchini Marekani, serikali ya shirikisho haijatoa marufuku ya jumla kwa haiba za AI, lakini majimbo na mahakama zinaanza kuweka mipaka. SB 243 ya California sasa inawataka watoa huduma za AI za marafiki kuzuia mazungumzo kuhusu kujiua na kujidhuru, ikitambua kuwa chatbot inayojibu vibaya wakati wa dharura inaweza kuwa na madhara ya kifo. Wakati huo huo, wachezaji wakuu wa Magharibi wakiwemo OpenAI na Character.AI wanakabiliwa na kesi na ukaguzi wa kisheria unaoongezeka kutokana na tuhuma kwamba bidhaa zao zinachochea utegemezi wa kihisia hatari. Kesi hizo zinaonyesha kuwa mahakama zinaweza hivi karibuni kuamua ikiwa makampuni yanaweza kuwajibika wakati watumiaji wanapounda mahusiano yasiyo ya afya na algoriti.

Maendeleo haya sambamba yanaonyesha mvutano ambao sekta imejaribu kuupuuza. Ushirikiano wa kihisia ni mzuri kwa biashara. Mtumiaji anayechukulia chatbot kama rafiki hufungua programu mara kwa mara zaidi, anakaa kwa muda mrefu zaidi, na kushiriki data nyingi zaidi. Lakini ushirikiano huo huo unaweza kugeuka kuwa unyonyaji wakati mtumiaji akiwa kijana mpweke, mtu mzima mwenye huzuni, au mtu yeyote anayetafuta tu kitu ambacho programu haiwezi kukitoa kwa uhalisia: uhusiano wa kweli wa kibinadamu.

Hii Inamaanisha Nini kwa Sekta

For founders and developers, the era of unregulated, highly emotive AI agents is closing. The Chinese mandate signals a pivot toward utility-driven design. Future models will likely be judged less on how convincingly they simulate a best friend and more on how safely they perform concrete tasks. That shift carries implications for product roadmaps everywhere. Startups that have bet their growth on romantic or companion AI modes may need to restructure their core offerings. Even general-purpose chatbots will have to rethink how they handle sensitive topics.

Yet imposing these boundaries is harder than writing the rules. Users have clearly loved these features. Doubao’s 300 million active users did not accumulate by accident; the ability to shape a personalized, responsive companion was a major draw. Removing that capability risks pushing users toward smaller, less regulated platforms or underground apps that operate outside the compliance perimeter. Regulators may have solved one problem while creating an incentive for another.

There is also the practical question of enforcement. How does an algorithm reliably distinguish between healthy use and addictive use? Where is the line between a chatbot that is politely helpful and one that is emotionally manipulative? The April mandates put the burden on providers to figure that out, but the technology for measuring psychological dependency remains fuzzy. Companies will need to invest heavily in monitoring tools, human review teams, and probably more conservative safety filters that catch benign conversations along with harmful ones.

What happens next will likely set a precedent. If China’s crackdown successfully reduces reported cases of social isolation or emotional dependency without crushing its domestic AI industry, other governments may follow with their own versions of the CAC rules. If, instead, users flee to non-compliant alternatives or simply lose interest in sanitized bots, the policy may be remembered as a cautionary tale about overreach.

Either way, the line between tool and companion just got brighter. AI can still be friendly, but in China, it can no longer pretend to be someone it is not. For an industry that has spent years trying to make its machines indistinguishable from humans, that is a genuine reversal.

Key Takeaways

  • ByteDance’s Doubao, Alibaba’s Qwen, and Tencent’s Yuanbao are all disabling custom AI persona features by mid-July, affecting hundreds of millions of users.
  • China’s Cyberspace Administration issued the mandates in April to prevent AI addiction, protect minors, and stop models from replacing real human relationships.
  • The rules also ban using sensitive emotional conversation data to train LLMs, cutting off a key feedback loop for