Week In Review: Perplexity makes a hefty bid for Google Chrome, the BBC says it will use GenAI audio for football summaries, and IPG & Aaru create predictive customer simulations.
Brands and Agencies
Dentsu & MOGL Partner for Sports Influencers
Dentsu is partnering with athlete influencer tech platform MOGL, giving Dentsu clients access to Name, Image, and Likeness (NIL) sponsorship opportunities. NIL refers to the rights athletes have to profit from the commercial use of their identity. MOGL is an AI-powered platform that uses first-party data to match brands with sports influencers. The tech enables Dentsu customers to create NIL campaigns while providing real-time performance reporting.
IPG Partners with Aaru to Build Consumer Simulations
Interpublic Group (IPG) announced a partnership with Aaru, an AI company focused on creating human simulations for campaign planning, creative testing, audience targeting, and more. The simulations will enable IPG clients to create synthetic audiences – that can emulate real human behaviour and forecast customer sentiment towards live events, ads, and corporate communications. IPG CEO Philippe Krakowsky, commented: “Combined with our exceptional Acxiom data asset and Interact platform, our partnership with Aaru will provide Interpublic and our clients with a distinct competitive advantage, enabling us to forecast campaign effectiveness and optimise creative executions with unprecedented speed and precision.”
S4 Capital Toys Possible Merger with MSQ
Sir Martin Sorrell’s S4 Capital was reportedly in talks with MSQ Partners for a potential merger, however yesterday MSQ ruled this out. The deal would have seen Sorrell’s S4 Capital buy MSQ, backed by One Equity Partners. S4 Capital shares climbed 3.8 percent after discussions of the merger arose. Agencies are looking for ways to bolster their offering amid the rise of AI. Omnicom is famously preparing to acquire fellow holding company Interpublic Group (IPG). However, MSQ said in a statement that even though discussions may have happened, there aren’t plans to have anymore.
AI Search Specialist Profound Raises $35m in Funding
Profound, an AI platform that focuses on brand visibility, raised $35m in funding in a round led by venture capital firm Sequoia Capital. The AI company, which launched last year, claims to be the first of its kind to seek brand visibility solutions in the AI search age. As more people turn to chatbots like ChatGPT or features like Google’s AI Overviews for search, publishers and brands fear their sites will be visited less often. These AI search tools often provide their own synopsis for an answer, removing the need to visit external sources. Profound’s technology works by letting brands know when AI bots visit their site and what kinds of queries related to the site are being made on AI search platforms. Profound CEO James Cadwallader claimed software firm Airbyte brought its visibility in ChatGPT from 9 percent to 26 percent in a week, ADWEEK reported.
Havas UK Unveils Tool to Track Brand Visibility in AI Models
Havas Media Network (HMN) UK announced the launch of Brand Insights AI, a new tool that gives brands intelligence into how they are appearing in large language models (LLMs). The tool scans six major AI platforms to give marketers insights into how their brand is showing up in AI chatbots. OpenAI’s ChatGPT, Anthropic’s Claude, Google’s Gemini, Perplexity’s Sonar, xAI’s Grok, and Chinese model DeepSeek are the models that can be scanned by Brand Insights AI. HMN clients can provide an overview of how brands are showing up across the six major models, including content analysis and a coverage and credibility score, giving brands better insight into the content driving AI rankings.
Media
BBC to Use GenAI Audio Summaries For Football Programmes
The BBC announced the launch of My Club Daily, a pilot AI service for football programmes on its BBC Sounds and BBC Sports platforms. The pilot will provide daily audio summaries for the Liverpool, Aston Villa, Newcastle United, Southampton, and Plymouth Argyle clubs. These audio updates will be generated using GenAI tools which “organise and reformat existing BBC articles about the club to produce a draft audio script.” The tool then creates the audio using an AI-generated voice to read the updates. The media company is using ChatGPT for the script generation and ElevenLabs to produce the synthetic voice. This pilot will run for four weeks from the beginning of the football season, and the updates will appear on BBC Sounds each day at 5pm.
James Cameron Calls GenAI ‘Genie In a Bottle’
James Cameron, the film director of blockbuster hits like Avatar and Titanic, has called GenAI the “next big wave in cinema technology” and “a genie that has been released from the bottle”. In an interview with ScreenDaily, the revered director called for the industry to address the “GenAI issue” and suggested it’s critical that it “remains an artistic tool and doesn’t replace artists”. Despite this, Cameron said he feels the tech can cut down production costs and enable film creators to make more imaginative movies.
YouTube To Test AI Age-Verification Systems
YouTube will begin testing an AI-powered age-verification system to safeguard minors from viewing inappropriate content. The AI system, to be tested in the US, guesses the age of the viewer regardless of the birthdate associated with the YouTube account. The system only operates, however, if a user is logged in. If the AI detects a user is under 18, YouTube will place privacy warnings on videos and restrict certain videos from being recommended. The video-sharing platform also does not show certain ads to users under 18.
Amazon Explains How its Nova Models Cite Sources
Amazon released a statement outlining how its Amazon Nova understanding models cite external sources in its responses. The company used a separate large language model (LLM) to evaluate Nova’s responses to different prompts. An example included asking the Nova model about Amazon shareholder letters. The statement attached the model’s output, showing the steps it took to answer the query. The model’s response came with numbered citations, including quotes the company confirmed were within the shareholder letter. Amazon released this statement to address the issue of untrustworthiness in AI models. The company explained how adding citations allows users to check if responses are truthful, as AI systems are prone to “hallucinations” – making up information.
Publishers See Decline in Referral Traffic from AI Overviews
Online publishers in the US reported an average 10 percent drop in referral traffic from Google compared to the year prior, according to survey results from trade association Digital Content Next (DCN). The 19 anonymous companies who participated in the survey stated that less readers are visiting sites after searching on Google, likely due to Google’s AI Overviews feature. The feature summarises a response to the user’s query, combining information from multiple sources. By having this overview at the top of the web page, users are less inclined to scroll and click on an article. DCN’s data revealed that the average year-on-year decrease in search referral traffic was higher for non-news publishers, at 14 percent. The Professional Publishers Association (PPA) in the UK also advocated for more transparency from Google in a statement to the UK’s Competition and Markets Authority. The industry body reported a 25 percent drop in article visibility for an automotive publisher. Another UK publisher reported a 25-30 percent lower clickthrough rate when AI Overviews are present.
Universal Pictures Adds Warning to AI Companies in Film Credits
Universal Pictures now includes legal warnings in the end credits of its films, cautioning AI companies against using its movies for AI training. The film production company prohibits the use of its films for training AI models. The warning states the movie “may not be used to train AI.” The company adds that it will carry out “civil liability and criminal prosecution” in the case of “unauthorised duplication, distribution, or exhibition.” The warning first appeared at the end of How To Train Your Dragon, released in June. It was also included in the credits of Jurassic World Rebirth and Bad Guys 2, which were released in July and August.
Tech
Musk Threatens to Sue Apple Over AI App Ranking
Elon Musk said his AI startup xAI will sue Apple for ranking the company poorly against AI competitors in its App Store. The controversial entrepreneur wrote on X that his AI company would take “immediate action” against what he believes is a breach of antitrust regulations. Musk also claims the app store favours X competitor OpenAI’s AI chatbot ChatGPT. “Apple is behaving in a manner that makes it impossible for any AI company besides OpenAI to reach #1 in the App Store, which is an unequivocal antitrust violation,” Musk wrote on X.
Cuttable Raises $4.5m in Seed Funding
Cuttable, a startup that uses AI agents to generate creative advertising on Facebook and Instagram, has raised $4.5m in seed funding. The Australian tech company has a total seed funding of $10m after 12 months of developing its product, and the funding round has been led by Square Peg. Other investors include Rampersand and Brand Fund. “We’ve been in deep build-mode for the past 12 months, now we are launching Cuttable to help growing brands make ads that drive sales through Facebook and Instagram,” says Sam Kroonenburg, Cuttable’s CEO.
Microsoft Launches Copilot 3D
Microsoft has launched Copilot 3D, a new tool designed to let users generate 3D models from 2D images. The tool can transform clean 2D JPG and PNG images into software, game engines, and augmented or virtual reality platforms. This news comes a week after Google announced the launch of its text-to-virtual world model Genie 3 – which will be used initially for simulations, gaming, and creating virtual universes in real-time.
SoundHound Launches Visual Understanding Platform
SoundHound, the voice and conversational AI-specialising company, launched Vision AI, an engine that adds vision capabilities to its existing voice AI platform. The engine combines the vision and hearing capabilities to create a platform that has a level of understanding better fit for human interactions. The technology uses cameras and speech recognition to help with tasks such as equipment troubleshooting, in-car assistance, and retail inventory intelligence. The company hopes to “redefine how humans interact with products and services offered and used by businesses,” said Kevyan Mohajer, CEO of SoundHound AI.
Google Launches Gemma 3 270M Model
Google introduced Gemma 3 270M, a multimodal model designed to efficiently carry out tasks. The model can be assigned to certain jobs and it then can carry out tasks like text classification and data extraction. Gemma 3 270M is suitable for “high-volume, well-defined” tasks like sentiment analysis, query routing, creative writing, and text processing. The model consumes less power and is Google’s “most power-efficient Gemma model,” Google says. Gemma 3 270M can run on less complex infrastructure and can complete tasks more quickly (in hours versus in days). The company explained it is not designed for “complex conversational use cases,” but more for following specific instructions.
Head of ChatGPT Considers Addition of Ads
Nick Turley, Head of ChatGPT at OpenAI, is thinking about bringing ads to the chatbot. Turley emphasised that he is “not humble enough to rule it out categorically” but that OpenAI must be “very thoughtful and tasteful” about ad integration into ChatGPT, he said on The Verge’s podcast Decoder. Bringing ads to the chatbot would be a way for the company to bring in some extra revenue. The company is also considering receiving a portion of revenue made from product purchases that come through ChatGPT. OpenAI chief executive officer Sam Altman has expressed reluctance in bringing ads to the platform, describing the move as a “last resort.” ChatGPT currently has over 700m users, many of which use the AI model for free – 20 million of its users were paying subscribers as of April, The Information reported.
Cohere Reaches $6.8 Billion Valuation
AI startup Cohere was valued at $6.8bn in its latest funding round, after securing $500m in funding. The startup’s valuation is up over $1bn from last year’s value of $5.5bn. Cohere, which specialises in AI models for enterprises, also recently recruited talent from other big-name companies. The startup hired Joelle Pineau, who previously led AI research at Meta, as its new chief AI officer. Cohere also appointed Francois Chadwick, former chief financial officer at Uber, as its own CFO.
Former Twitter CEO Announces $30 Million AI Startup
Parag Agrawal, former CEO of Twitter (now known as X), has created his own AI startup, Parallel Web Systems Inc. The AI platform is designed for conducting research and has received $30m in funding, The Information reported. It uses AI applications to collect data from the internet and produce responses. The system has eight different “research engines” with designated speeds and information-gathering abilities. The system is useful for a variety of research tasks, including automating workflows and coding assistance, the company explained in a statement. The company claims Parallel Web Systems is “the only AI system to outperform both humans and leading AI models like GPT-5 on the most rigorous benchmarks for deep web research.”
Blue Lucy Unveils Media Workflow Tool
Media management and workflow automation company Blue Lucy has launched a new capability into its product ‘BLAM’ that enables producers, distributors and content creators to integrate numerous LLMs into their media workflows. This approach gives marketers control over performance, cost and data privacy. The company is also set to launch ‘Lucia’, a native AI assistant that helps streamline workflows.
Number of the Week
$34.5 billion. That’s how much the US-born AI startup Perplexity is reportedly trying to buy Google Chrome for. Revealed on Tuesday by The Wall Street Journal, the AI answer engine has made a bid for the world’s most popular search browser with roughly 3 billion users. This move signifies the growing trend of AI companies launching their own AI browsers – Perplexity unveiled its own browser ‘Comet’ last month. AI is upheaving online search, changing the way users look for information and search for products. Some investors have called the bid a ‘stunt’, saying Chrome is worth much more than Perplexity’s offer.
