2026国际消费电子展 (CES) ：从媒体采买到决策影响研究报告

From Media Buying toDecision Influence How AI-Powered Devices Are Rewriting Advertising,Commerce, and the Role of Agencies Foreword Every year’s, CES unveils innovations that can seem incremental in isolation, such as smarter TVs, moreconnected cars, lighter wearables, but when viewed together they point to a far more profound shift:the emergence of a new operating system for consumer decision-making. AI is no longer confined to apps or platforms; it is being embedded directly into the devices thatshape daily life, observing context, interpreting intent and increasingly mediating choices in real time. This matters because media, advertising and commerce have always hinged on one criticalquestion—when to show up—and AI is now redefining that moment. As traditional, channel-basedmedia models strain under fragmentation, privacy constraints and scarce attention, AI-powereddevices are generating a continuous stream of signals—not cookies, but real-time indicators ofbehaviour and intent—that automate decisions across environments. Written for brand leaders, media strategists and agencies navigating a post-cookie world, this reportexplains why the next era of media will be won by understanding how AI decides what getssuggested, chosen and trusted—and what that means for staying relevant over the next decade. Ricky ChopraGlobal Head of Innovation and AI, Zenith Media Signal Ownership: How AI-Powered TVs Are TurningTelevision into a Programmatic Marketplace At this years CES, LG introduced the next generation of its AI-powered television platform,embedding on-device AI directly into the operating system. Rather than functioning purelyas a content navigation tool, this AI layer learns from viewing behaviour, interactionpatterns, and contextual usage — processed locally and anonymised.This represents a fundamental shift in the role of the television. The TV is no longer just a screen for content delivery; it is becoming a signal-generatingdevice. As broadcast viewing continues to decline and on-demand consumption becomesthe default, intelligence increasingly sits at the device level — not the channel level. From Broadcast Reach to Signal-Based Addressability Traditional TV buying was built onapproximation: •Programmes as proxies for audiences•Dayparts as proxies for behaviour•Reach as a proxy for influence As television moves from broadcast to on-demand, these proxies will begin to break down.What may replace them is not simply “addressable TV”, but something more powerful: AI-derived household signals. What LG Now Owns: Signals,Not Inventory The Shift to Programmatic Television Logic With AI embedded at the device level, LGeffectively owns a new layer of value:anonymised, predictive household signals. This evolution takes television decisivelycloser to programmatic buying mechanics.The logic becomes familiar: These are not personal identifiers.They are probabilistic insights such as: •Signals are generated by the device•Signals are anonymised and aggregated•Advertisers decide which signals align withtheir objectives•Media activates when those signals arepresent •Upgrade intent•Price sensitivity•Lifestyle orientation•Category openness•Decision-making velocity For media buyers, this is a meaningful shift. Insteadof buying TV inventory tied to content, advertiserscan begin to buy access to signals that indicaterelevance. Rather than planning against schedules orshows, media buyers plan against conditions.For example: •When a household exhibits earlyautomotive consideration signals•When entertainment behaviour shiftstowards premium experiences•When price sensitivity increases•When lifestyle patterns suggest a new lifestage In this model, LG does not primarily sellimpressions. LG sells the opportunityto bid against inferredintent. This is television buying that behaves likedigital — but with the scale, attention, andhousehold truth of TV. What This Means for Media Buying Teams As AI-powered TVs scale, media buying will increasingly resemble signal trading rather than inventoryplanning. Key changes include: •Strategy shifts from channel selection to signal selection•Optimisation shifts from reach curves to signal performance•Measurement shifts from exposure to probability of outcome This also reframes the role of agencies. Value no longer comes from access to inventory, but from: •Interpreting signal quality•Deciding which signals matter•Connecting signal activation to real-world outcomes The Consumer Perspective This model only works if it delivers value to consumers. Because signals are: •Anonymised•Processed on-device•Used to reduce irrelevant advertising•Consumers experience:•Fewer ads•More relevance•Less repetition•Better alignment with real needs Signal ownership, when governedresponsibly, becomes a relevance engine— not a surveillance model. Our Prediction: Television Becomes aSignal Exchange Why This Matters Now CES announcements confirm thatintelligence is moving onto thedevi