Welcome to Week In Review
From new tools transforming creative work, to campaigns reshaped by automation, we track the biggest shifts at the intersection of AI, media, and marketing. This week, Yahoo launches AI answer engine, AdCP’s sales agents move to Prebid, and the UK Competition and Markets Authority makes moves to block publisher content from being used in AI Overviews.
Top Stories of the Week
Yahoo Launches AI Search Tool ‘Scout’ to Challenge Google and Perplexity
Yahoo has launched Scout, an AI-powered “answer engine” designed to deliver conversational responses rather than traditional search results. Built on Anthropic’s Claude model and supported by Microsoft’s Bing grounding API, Scout draws on Yahoo’s data, content and user profiles to power search across news, finance, shopping and live information, marking Yahoo’s formal push into the AI search race.
Why it matters: Scout signals that AI search is no longer a two-horse race. As more platforms become answer engines, visibility, attribution and influence will fragment beyond Google. Brands and publishers must optimise for how AI systems source, summarise and prioritise information.
AdCP’s Sales Agents Move to Prebid
Prebid, the open source framework, has announced it will take ownership of open-source code for a publisher-side AI sales agent built on the Ad Context Protocol (AdCP), aiming to help publishers participate in emerging AI-to-AI programmatic trading. The new Prebid Sales Agent, available via GitHub, is designed as a reference implementation to prevent fragmentation and ensure publishers aren’t sidelined as agentic advertising frameworks evolve.
Why it matters: Programmatic advertising is on the verge of shifting from human-led trading to AI-to-AI negotiation. By anchoring sell-side agents in open-source infrastructure, Prebid gives publishers a way to stay competitive without becoming dependent on proprietary platforms.
UK Regulator Moves to Let Publishers Block Google’s AI Overviews
The UK Competition and Markets Authority (CMA) has proposed allowing publishers to block Google’s AI Overviews from scraping their content, signalling a shift toward giving news organisations greater control over how their work is used in AI-generated search summaries. The move follows sharp traffic declines blamed on AI overviews, though critics argue the proposals fall short of compensating publishers for lost revenue and influence.
Why it matters: The move from the CMA marks a potential rebalancing of power in AI search. If publishers can block AI Overviews, brands may lose a key discovery surface while news outlets regain leverage over distribution. It raises urgent questions about visibility, referral traffic and compensation in AI-mediated search, and whether regulation can realistically protect media economics as user behaviour shifts toward instant, summary-based answers.
Quote of the Week
Brands & Agencies
Tourism Fiji Partners with Havas for AI-Powered Campaign
Tourism Fiji has launched a new iteration of its Where Happiness Comes Naturally campaign, inviting travellers to “Be Fiji.” Developed with Havas, the work targets everyday “Un-Fiji” moments using contextual media and Converged.AI, Havas’ media planning and activation platform, positioning Fiji as an antidote to modern stress while pairing emotional storytelling with tactical travel offers.
Publicis Bets on Privacy-First AI With Decentriq Data Partnership
Publicis Groupe has partnered with Decentriq to build privacy-first AI and data collaboration capabilities for marketing and media. Using zero-copy, encrypted data infrastructure, the alliance allows Publicis and its partners to run analytics and AI agents on sensitive first-party data without sharing or exposing it – strengthening trust, compliance and outcome measurement as AI adoption accelerates.
PVH Brings OpenAI for Calvin Klein and Tommy Hilfiger
PVH Corp has signed a partnership with OpenAI to integrate its enterprise APIs across the company’s global operations, supporting teams working on consumer engagement, product design, supply chain and retail. The fashion group, which owns Calvin Klein and Tommy Hilfiger, said the collaboration will focus on building bespoke internal workflows and accelerating data-driven decision-making through a test-and-learn approach deployed at scale.
Dentsu Launches Generative Audiences to Rebuild Media Targeting
Dentsu has launched Generative Audiences in the UK, an AI-powered audience intelligence and activation capability designed to move media planning beyond traditional segments. Combining deterministic people-based data with AI-driven signals, the solution creates privacy-conscious, synthetic personas calibrated to real-world behaviour, allowing marketers to generate insights, test scenarios and activate media with greater contextual relevance and precision.
Fanatics Collectibles Uses Adobe GenAI to Turn Live TV Into Digital Cards
Fanatics Collectibles is using Adobe’s Firefly GenAI tools to dramatically speed up production of digital trading cards, cutting turnaround times from weeks to hours and reclaiming more than 60 percent of staff production time. The approach allows Topps Now to release digital collectibles within 24 hours of live games, capitalising on peak fan interest and boosting real-time engagement around major sports moments.
Media
Australian Journalism ‘Invisible’ in AI News Summaries, Research Finds
Research from the University of Sydney has found Australian journalism is largely absent from AI-generated news summaries produced by Microsoft Copilot, which overwhelmingly prioritise US and European outlets. The study warns this imbalance risks accelerating news deserts, weakening local media sustainability and undermining democratic trust, as AI tools increasingly mediate how audiences access news.
Anthropic Revealed to Have Scanned Millions of Books to Train AI Models
Anthropic has been revealed to have spent tens of millions of dollars purchasing and scanning millions of physical books to train its AI models, according to court documents reported by The Washington Post. The disclosure highlights how AI companies are turning to large-scale physical data collection to secure high-quality training material, intensifying debates around copyright, fair use and the future economics of creative content.
Thinktank Calls for ‘Nutrition Labels’ and Mandatory Payments to Publishers
The Institute for Public Policy Research has urged the UK government to require transparency labels on AI-generated news and introduce a licensing regime forcing AI companies to pay publishers for content use. Warning that AI platforms are becoming powerful new gatekeepers, the thinktank said intervention is needed to protect media plurality, trust and the long-term sustainability of independent journalism as AI increasingly mediates news consumption.
Japan’s Anime and Manga Industry Hit by $38bn Piracy Loss
Japan’s Ministry of Economy, Trade and Industry has reported that anime and manga piracy cost the country an estimated $38 billion in 2025, nearly triple the losses recorded in 2022. The findings have prompted METI to outline new countermeasures, including tougher litigation frameworks, action against GenAI-enabled infringement, and stronger enforcement against counterfeit character goods.
Tech
Apple Set to Debut Gemini-Powered Siri in February
Apple is expected to unveil a new version of Siri powered by Google’s Gemini models in February, according to Bloomberg. The update is set to enable more advanced, task-oriented assistance using personal and on-screen data, marking Apple’s first tangible output from its Google AI partnership ahead of a more conversational Siri upgrade planned for WWDC.
Pinterest Cuts Up to 15% of Staff as It Redirects Resources Toward AI
Pinterest has announced plans to lay off fewer than 15 percent of its workforce as it reduces office space and shifts investment toward artificial intelligence initiatives. The cuts, expected to affect around 700 employees, come as the company accelerates development of AI-powered products such as Pinterest Assistant and personalised boards, while managing restructuring costs of up to $45 million.
Google Brings Gemini Deeper Into Chrome With Agentic AI
Google has announced a new wave of Gemini-powered features for Chrome, including image generation, personalised intelligence and Auto browse, an agentic AI tool that can complete tasks across the web from a single prompt. The updates embed Gemini directly into the browser interface, underscoring Chrome’s role as Google’s frontline in the race to build a universal, action-taking AI assistant.
Shotstable Debuts AI Filmmaking Tool
Shotstable has released an alpha demo of its AI-powered filmmaking platform, introducing an AI Shot Table designed to help creators plan and generate films with greater control. The tool focuses on maintaining character and scene consistency, enabling batch generation and consolidating fragmented AI workflows into a single, shot-based interface aimed at accelerating AI film production.
Zefr Gets MRC Accreditation for YouTube Viewability Measurement
Zefr, an AI content classification tool, has received Media Rating Council accreditation for its independent third-party processing and reporting of YouTube viewability through Google’s Ads Data Hub. Following an external audit, the approval validates Zefr’s measurement standards and coincides with the launch of free, independent YouTube viewability reporting for advertisers, aimed at improving transparency and accountability in video ad measurement.
Palazzo Launches AI Studio to Automate Lifestyle Imagery
Palazzo has introduced Palazzo Studio, an AI-powered creative platform that generates photorealistic lifestyle scenes from standard product images for furniture and home brands. Designed for e-commerce and merchandising teams, the tool helps fill PDP and PLP imagery gaps, maintain visual consistency at scale, and reduce production costs, reflecting how AI is reshaping product visualisation and conversion-focused content.
Snap Unveils Specs Smartglasses as a New AI-First Computing Platform
Snap has outlined its vision for Specs, a new generation of smartglasses designed to blend digital experiences directly into the physical world through see-through lenses, hand and voice controls, and a built-in AI “Intelligence System.” Positioned as an alternative to screen-based computing, Specs aim to keep users present while using AI to anticipate needs, support shared experiences, and reduce reliance on physical devices and materials.
Start-Up of the Week
Limy has emerged from stealth with $10 million, in funding led by Flybridge, to build infrastructure for brand performance in the agentic web. The startup focuses on analysing AI agent and bot behaviour – rather than user prompts – to help brands improve discoverability, engagement and revenue as AI-driven agents increasingly mediate how information is accessed and acted upon online.
Number of the Week
Meta reported full-year revenue of $200.97 billion in 2025, up 22 percent year-on-year, marking a major milestone for the company. Advertising accounted for nearly all growth, with rising impressions, improved pricing and AI-driven delivery across Facebook and Instagram offsetting losses in Reality Labs and reinforcing Meta’s near-term reliance on ad performance as it invests heavily in AI infrastructure.
