Week In Review: BMW Group chooses Dentsu for AI advertising, Havas reports Q2 growth thanks to AI efforts in the US, and YouTube named the top receiver of AI search traffic.
Brands and Agencies
BMW Chooses Dentsu for AI Advertising
BMW Group announced it would be partnering with Japan-headquartered holding company Dentsu to boost its advertising strategy in Europe. The partnership will commence in January 2026, and focus on bolstering BMW and its brands with the use of AI and centralised data management, to make campaigns more efficient and targeted. Uwe Dreher, Vice President Marketing BMW Europe, said the car manufacturer’s goal is to use AI for a “data-driven and targeted multi-channel and multiple market planning approach” to “increase efficiency and impact”.
WPP OpenDoor Unveils AI Emotional Recommendations
WPP OpenDoor, the team for Amazon’s account at WPP, has unveiled a campaign in India for Prime Video using AI. The campaign develops an emotion-sensing layer within Prime Video’s digital ad space so that users interacting with ad banners get tailored recommendations based on their mood. Deepa Jatkar, India Lead at WPP OpenDoor, said: “In today’s dynamic digital landscape, true personalisation transcends data points; it’s about understanding the human experience. With Prime Video, we saw an unparalleled opportunity to innovate beyond behavioural targeting and connect with audiences at a deeper, emotional level.”
Havas Reports Overall Q2 Growth
Havas reported strong revenue growth in the second quarter of this year with 2.6 percent organic growth, up from 2.1 percent. This is despite some challenges the holding company has faced in the APAC and Africa regions, which had a 4.9 percent decline. Leading the way was North America, where there was organic growth of 4.6 percent in Q2 and a 3.9 percent in the first half of the year. Havas said growth in the States was supported by investments in AI, including the holding company’s own Converged.AI – a media planning, production and personalisation tool.
WPP Media and Criteo Partner for AI Enabled Advertising
WPP Media, the holding company’s ad buying agency, and media platform Criteo have partnered, uniting WPP’s AI platform Open Intelligence with Criteo’s supply-side platform (SSP), commerce grid, and data signals. Open Intelligence, launched in June this year, is a AI “large marketing model” that can predict audience behaviour and market performance. According to a statement, a combination of Open Intelligence, relationships with supply partners, and Criteo’s real-time commerce signals, will enable advertisers to reach consumers at scale.
Cheil UK and AI Startup Spotlight Partner
Integrated marketing and advertising agency Cheil UK announced a strategic partnership this week with Spotlight, an AI startup focused on brand visibility and performance in the age of AI. The partnership will enable Cheil to build proprietary tools designed to help clients continue to be discovered by consumers and rank highly in AI search. The collaboration is part of wider AI investment plans from the agency and will sit alongside other AI tools. “This is not about bolting AI onto old ways of working,” said Chris Camacho, CEO at Cheil UK. “It’s about building the future from the inside out. Our clients need tools designed for the way the world is changing. Spotlight shares our urgency and ambition. This partnership marks a new phase in AI-led marketing development, helping brands take a leading role as search and discovery models continue to evolve.”
Agentic AI is Most Significant Marketing Trend, Says McKinsey
McKinsey’s Technology Trend Outlook 2025 report found that agentic AI – which is the use of AI assistants that can perform tasks and reason autonomously – is the most significant emerging trend for marketers right now. So far, agentic AI has helped many marketing teams automate tasks like campaign optimisation, planning, audience targeting, and performance analysis, without needing continuous oversight. “Agentic AI moves AI from a passive tool to an active collaborator with enterprise workflows,” said Delphine Nain Zurkiya, senior partner at McKinsey Boston.
Media
Google Launches ‘AI Mode’ Search Feature in UK
Google launched ‘AI Mode’ in the UK – a new search feature that lets Google users ask questions within an AI-powered chatbot. This differs from Google’s already available ‘AI Overviews’ which summarises search query results at the top of Chrome’s web browser page. However, the new offering adds to growing concerns for online publishers, and online retailers, who rely heavily on search traffic as part of their business model.
Amazon and The New York Times Establish $25 Million Licensing Deal
Amazon will pay The New York Times (NYT) $20-25m a year for content, The Wall Street Journal (WSJ) reported. This deal allows Amazon to source NYT content to train and improve its models. The deal was originally announced in May, however exact details weren’t available. It is the first AI licensing deal for The New York Times and for Amazon. Similar agreements have been made between AI developers and publishers. OpenAI holds agreements with publishers such as News Corp, the Associated Press, and the Financial Times. These deals allow developers more access to information on which models can be trained, and they increase the visibility of publisher content in AI search engines.
Reddit Predicts AI-Driven Ad Revenue Boost
In its third-quarter revenue forecasts, Reddit estimated a revenue increase to $535-545m. This revenue, around $70m above average analyst predictions, also exceeds its second quarter revenue of $500m – a 78 percent increase from the previous quarter. The company’s shares rose over 16 percent. The social media platform cited its AI ad tools, which are based on user conversations, and success in AI search traffic as reasons for the boost. It also includes an AI-driven translation feature which helped expand its user base. Reddit’s daily active users increased 21 percent to 110.4mn users in the most recent quarter. The company also stated that Google’s ‘AI Overviews’ are helping boost site traffic, as Reddit is a frequently cited source by the AI search tool.
Adobe Invites Users to Finish a Film Project with AI
Adobe released a social campaign, titled ‘The Unfinished Film,’ inviting creatives to collaborate on a project using the company’s Firefly AI tools. Adobe partnered with director Sam Finn to create this project. The Unfinished Film features colourful imagery, including a person running through a pixelated forest and a boat drifting through a museum-like space. The project beckons directors, artists, designers, and other creatives to fill in the missing elements. “It used to take years of learning–and a lot of money–to make a film,” Finn says. “Now anyone with a vision can start.”
Creative Groups Express Concern Over EU AI Code of Practice
A group of authors, publishers, producers, and other European rightsholder organisations released a statement expressing their “dissatisfaction” with the European Union’s (EU) General-Purpose AI (GPAI) Code of Practice. The GPAI Code of Practice, released on July 10 2025, is a set of guidelines to help AI companies comply with the EU AI Act obligations. It includes three pillars – transparency, copyright, and safety and security. However, creative organisations from across the region have said the GPAI Code of Practice doesn’t go far enough in protecting creative sectors in the region from having their work scraped by AI companies to train AI models. Google and Microsoft announced they will sign the EU Code of Practice, while Meta declined to sign on the grounds of it stifling AI innovation.
BBC News Appoints AI, Innovation and Growth Director
BBC News has appointed a new AI, Innovation and Growth Director, Anjali Kapoor, to oversee AI adoption for the broadcaster’s news division. Kapoor has previously worked at Meta, where she was Director for Media Partnerships in the Asia-Pacific region, and has also had roles at The Globe and Mail, Yahoo!, and Bloomberg Media. In a statement, the BBC said Kapoor will help lead the broadcaster’s “off-platform growth and focus on optimising BBC audiences and digital strategy for growth in the UK and globally.”
Runway’s AI Film Festival to Screen in IMAX Cinemas
AI video-generation platform Runway is partnering with IMAX to screen movies made for its 2025 AI Film Festival. Finalists from the film festival will have their creations shown across IMAX cinemas in the US, including in Atlanta, Boston, Chicago, Dallas, Denver, Los Angeles, New York City, San Francisco, Seattle, and Washington DC. The short films, ranging between 2 and 10 minutes, will run in these cinemas from August 17 to August 20, but will be available to watch online following the IMAX screenings. “The quality, variety and storytelling of these films deserves a premium viewing experience,” said Cristóbal Valenzuela, co-founder and CEO of Runway.
Reach Newspapers to Add Paywalls Amid AI Concerns
Reach PLC, the publisher behind the Daily Mirror and the Daily Express newspapers, will trial paywalls on its digital content, The Telegraph reported. This decision comes as publishers face lower levels of website traffic and readership because of AI. The publisher will pilot digital subscriptions for its news outlets later this year. This includes for its regional publications such as Manchester Evening News and Liverpool Echo. Reach will experiment with tiered subscription models on its articles. The company said that while readers will have to pay to read some articles, not all articles will have a paywall.
Tech
Google Indexes ChatGPT Conversations
The Fast Company reported that Google has been indexing conversations from ChatGPT that users have sent to colleagues, friends and family – putting private exchanges into search results. It was found that if users click ‘Share’ to send conversations within the ChatGPT interface, it is automatically indexed in Google’s search results. Despite the identity of users being anonymous, chat content – which has included personal interactions between users and the AI chatbot around mental health and physical abuse – risks searchers being identified.
Google Cloud Launches Veo 3 on Vertex AI
Google updated Veo 3, its video generation model, to work faster. ‘Veo 3’ can generate short videos from images and text and ‘Veo 3 Fast’ supports faster conversion from text-to-video. The updated ‘Veo 3 Fast’ allows users to create short, AI-generated video clips, at a quicker rate than before. This means it can more quickly generate advertisement iterations or demos, expediting production processes. The upgraded model supports generating videos in 1080p HD, with audio generation features including lip-synching and sound effects. Google Cloud announced it will be launching Veo 3 and Veo 3 Fast through Vertex AI – Google’s platform for building, training, and deploying models. The tech giant said it has seen massive global adoption of Veo, especially among enterprise customers, with over 70 million videos created since May.
Hightouch Announces Investment from Snowflake
Hightouch, a data and agentic AI platform, announced new investment from Snowflake Ventures and Capital One Ventures. The investment extends Hightouch’s Series C round, indicating support for the growing agentic space. One of Hightouch’s primary offerings is its tool ‘AI Decisioning’, that lets marketers deploy agents that can constantly experiment and improve targeting, iterating campaigns continuously. “This investment is a major vote of confidence in our vision for AI-powered marketing,” said Tejas Manohar, Co-CEO and Co-Founder of Hightouch. “We’ve seen incredible traction with AI Decisioning, and this investment enables us to scale faster, deliver even more value to customers, and continue reimagining how businesses personalise experiences through data and AI.”
YouTube Is Top Receiver of AI Search Traffic
YouTube receives the most AI-driven search traffic compared with other websites, according to data from software company Similarweb. The report tracked the Top 50 Sites Getting Traffic From AI Chatbots and found that YouTube, the video sharing website owned by Google, has garnered three-times as much traffic than websites Facebook (10.9m) and Wikipedia (10.1m) – which came in second and third place – with 37.9m in May 2025.
Spotify Hints at Upcoming AI Features
Spotify plans to expand voice AI capabilities in the future, according to executives from the company. In a call with investors, Spotify Chief Product and Technology Officer Gustav Söderström shared that the music sharing platform will become more interactive for its users. “You can already write to Spotify, talk to Spotify. You’re just going to see that expand.” Söderström said. Spotify already has an AI-powered “DJ” feature, which allows users to vocally request songs or genres. The AI tool plays music based on the user’s listening habits. Söderström shared that these user-DJ interactions have helped the company gather more data on what phrases, ideas, and songs may connect well with each other.
Google Adds AI Summaries for Shopping Assistance
Google announced an update to its Chrome web browser that provides AI-generated review summaries to provide shopping assistance. The feature appears as a pop-up icon on the browser, showing users a summary of previous reviews on the product or store the user is viewing. It provides information on aspects such as pricing, product quality, and customer service experiences. The tool generates summaries by gathering information from customer review platforms such as Trustpilot, TurnTo, and Yotpo. The feature will be available on Chrome on desktops.
Microsoft Introduces ‘Copilot Mode’
Microsoft added ‘Copilot Mode’ to its Edge browser. This AI feature helps users complete tasks with minimal prompting. This includes gathering and organising information from multiple tabs the user has open without the user switching between them. Copilot Mode will comprise search, web navigation, and chat into a single entity. Voice navigation features will also soon be added to Copilot. Copilot will only access content when it is approved by the user and will indicate when it is active, Microsoft said. Users can opt out of the feature at any time.Copilot will soon be able to be granted access to user history to assist it in completing tasks like booking events or setting to-do lists. The feature will be available for free in Windows and Mac PC devices.
Adobe Rolls Out AI Image Editing Upgrades to Photoshop
Adobe is adding more AI image editing capabilities to Photoshop – an image-upscaling feature, an object blending tool, and an upgraded object removal tool. The editing features are powered by the company’s Firefly models. The “Generative Upscale” feature can enhance an image by up to 8 megapixels, improving the quality and sharpness. “Project Perfect Blend” offers colour-correcting, shadows, and lighting adjustments. This tool can alter the lighting of an object added on top of an image to blend them together. Its third new feature, an upgraded object removal tool, uses AI to more seamlessly fill in image backgrounds after removing an element from the image. Adobe allows users to select which GenAI Firefly model they would like to deploy while using AI features.
Amazon Restricts Access of Other AI Shopping Tools
Amazon reportedly altered its code to prevent external AI shopping assistants like Google from accessing its website, The Information reported. The e-commerce giant is restricting AI tools from other companies including OpenAI, Perplexity, and Anthropic. This restriction extends to Amazon’s product visibility in other chatbots. The report said a search for Amazon products on bots like ChatGPT returned no Amazon product results. This move from Amazon is an attempt to maintain site visits, as the rise of AI chatbot use for shopping puts online retailer traffic at more of a risk.
Anthropic Models are Preferred by Enterprises
Anthropic accounts for 32 percent of enterprise large language model (LLM) share, according to a Menlo Ventures report. This is a big jump from 2023, when the company held only 12 percent of market share by usage. OpenAI holds 25 percent of market share by usage, decreasing by half since 2023 (50 percent). Anthropic’s coding models are extremely sought after by enterprises, with these models encapsulating 42 percent of enterprise market share, also beating out OpenAI’s 21 percent. Anthropic’s Claude 3.5 Sonnet model, designed for advanced coding and reasoning, led to the boost in adoption, according to the report. The company released the Claude 3.7 Sonnet model in February, which the company says offers a higher level of intelligence and the ability to solve complex problems.
Microsoft Sees Revenue Boost, Driven by AI
Microsoft reported an 18 percent revenue increase from last year to $76.4bn. The company released its financial results from the three month quarter ending on June 30. Microsoft attributed this revenue increase to the success of its cloud and AI products. “Cloud and AI is the driving force of business transformation across every industry and sector,” said Satya Nadella, Chairman and Chief Executive Officer at Microsoft.” This quarter, Microsoft added multiple upgrades to its Copilot AI platform. This included adding a researching agent, a data analysing agent, an image-generation tool, and an audio overview tool. The company recently became the second company –the other being chipmaker Nvidia–to reach a $4tn valuation.
Meta Reports Revenue Boost Due to AI-Backed Ads
Meta Platforms, which owns Instagram, Facebook and WhatsApp, reported stronger than expected revenue in the third quarter of 2025. The company said revenue from April-June rose 22 percent from last year to $47.5bn (£35.86bn), and profits rose 36 percent to $18.3bn. Analysts have pinned the use of AI powering the tech giant’s ad business as the reason for the hiked results, which climbed 11 percent in extended trading. In June this year, the Wall Street Journal reported that Meta would be enabling AI ads, where advertisers can generate and target ads using AI, with full implementation by the end of 2026.
Number of the Week
$170 billion. That’s how much AI start-up Anthropic aims to reach in its valuation, more than doubling current figures. The US-based start-up is seeking to dramatically increase its worth in a new funding round led in the Middle East. Bloomberg reported this week that the ethically-focused parent company of the large language model (LLM) Claude is apparently in talks with the Qatar Investment Fund (QIA) to take part in a hefty funding round. While Anthropic has been viewed as the LLM prioritising safety and ethics, the company’s CEO Dario Amodei reportedly told employees in a memo that to stay competitive it can’t afford to ignore the “truly giant amount” of capital in the Middle East.
