The news: A CBS investigation discovered hundreds of deepfake ads on Meta platforms promoting “nudify” apps that create sexually explicit content based on images of real people. The analysis of Meta’s ad library found at minimum hundreds of deepfake ads across Facebook, Instagram, Threads, Facebook Messenger, and Meta Audience Network. Our take: The rise of deepfakes on major platforms like Meta emphasizes AI’s potential to erode consumer trust and raise brand safety risks—forcing advertisers to navigate a growing gap between innovation and lagging safeguards.
The news: Generative AI (genAI) has become standard across US enterprises—95% of companies report using it to some extent, up from 83% a year ago, per Bain & Co—but wider enterprise adoption is hitting roadblocks. A lack of robust governance and the need for continuous security validation are getting in the way. Our take: To escape limbo, enterprises must shift from experimentation to disciplined execution. That means building AI governance into the foundation—not as an afterthought. Security, transparency, and trust must be embedded into every AI deployment. Businesses shouldn’t just see AI as a plug-and-play solution without vetting it and aligning it with desired outcomes. For marketers, campaigns built on shaky AI foundations risk brand reputation, compliance failures, and consumer mistrust.
The news: Apple’s highly anticipated AI enhancements, particularly for Siri, remain unfinished. During WWDC 2025, SVP Craig Federighi confirmed delays, stating Apple needs “more time to reach a high-quality bar.” No major voice assistant upgrades were announced. Apple’s most relevant AI move wasn’t a product—it was a warning: Ahead of its event, Apple published a research paper arguing that top models like OpenAI’s GPT-4o, Anthropic’s Claude 3.7, and Google’s Gemini don’t truly “reason.” Instead, they create an “illusion of thinking.” Our take: Apple is hedging its AI bets by being cautious with core offerings like Siri while quietly enabling developers with on-device LLMs and privacy-first tools. Instead of overpromising, Apple is pointing out potential problems with the latest AI models while exercising restraint.
The news: Mark Read will exit as CEO of WPP at the end of 2025, concluding a three-decade run shaped by AI investment and structural overhauls. While Read launched tools like WPP Open and pushed to streamline operations, the company still posted a 1% organic revenue decline in 2024 and hit a four-year stock price low. Our take: Read’s departure marks a critical inflection point for WPP and the broader agency model. With 56.1% of agency leaders naming inefficiency as their top issue, the next CEO will need to go beyond tech implementation and deliver meaningful workflow clarity and cost discipline—fast.
On today’s podcast episode, we discuss how Americans’ feelings towards AI have changed this year, the gaps in concern between AI experts and the general public, and the best ways to get started with AI. Join Senior Director of Podcasts and host Marcus Johnson, Analyst Grace Harmon, and Senior Vice President of Media Content and Strategy Henry Powderly. Listen everywhere and watch on YouTube and Spotify.
A large majority of US consumers are somewhat (26%), very (32%), or extremely (34%) concerned about AI spreading misinformation, according to an August 2024 survey from the Pew Research Center.
The news: Walmart rolled out Sparky, its generative AI (genAI) assistant, to all Walmart app users this week—a preliminary step that puts it closer to achieving its agentic ambitions. Our take: By broadening Sparky’s capabilities, Walmart is trying to position itself not only as a shopping destination, but also as a place where consumers can go when they need everyday life advice or information—such as how to fix a leaky faucet or help with event planning. Whether the retailer succeeds will depend on how well Sparky works, and whether it can convince shoppers to overcome their current skepticism of AI tools.
The news: Meta is in talks to invest upwards of $10 billion in Scale AI, a data labeling startup. The deal would be Meta’s biggest ever external AI investment and could help it position its Llama large language model (LLM) as an industry standard, per Bloomberg. Scale AI has already partnered with Meta to develop Defense Llama, an LLM designed for military use that’s built on Llama 3, and also works with Meta competitors like Microsoft and OpenAI. Our take: Meta’s massive investment could draw antitrust scrutiny in an era of acqui-hires. The outcome of active probes in Big Tech partnerships could influence regulatory action, especially if this investment contains any exclusivity that limits model training resources for other companies.
The news: IT leaders are increasingly concerned about unauthorized employee use of AI and its risk to company security and compliance. 90% are concerned about “shadow AI,” or employees adopting AI tools without IT team approval, per Komprise’s IT Survey: AI, Data & Enterprise Risk. 13% of companies said genAI has harmed their finances, customers, or reputation—proof that AI’s risks aren’t just hypothetical. Our take: Companies should pair data management and protection of sensitive data with worker training. Giving employees access to tools they’ll actually use and keeping them in the loop on AI plans could help prevent the use of unauthorized tools and data leaks, foster trust, and deter sabotage of genAI initiatives.
The news: This week’s Apple Worldwide Developers Conference (WWDC) will be a critical opportunity for Apple to define its AI transformation after a year of missteps, unfulfilled promises, and user fallout. Our take: Apple must convince users and developers that its platform is where meaningful AI happens. Leaning solely on OS and service updates won’t cut it, and ignoring its AI roadmap risks slowing iPhone and Mac upgrade cycles. The pressure is mounting. Samsung and Google are packing AI into their next phones, and 1 in 5 iPhone users say AI features could drive their next smartphone upgrade, per CNET.
The news: In-car voice commerce has the potential to unlock a $35 billion annual opportunity for automakers, according to new research by in-car voice technology provider SoundHound AI. The hands-free tech integrates voice ordering, payments, and navigation directly into vehicles—transforming them into mobile commerce hubs that users are already familiar with. Key takeaway: Marketers and advertisers should prepare for a shift in automotive user interface by integrating voice-first campaigns into connected car ecosystems. Opportunities include forging partnerships with automakers and service providers for branded voice experiences, sponsored suggestions, and frictionless ordering while prioritizing transparency to satisfy safety regulators.
The news: Google launched “Portraits” on Thursday, a new experiment that creates AI-generated versions of influencers that offer users notes and advice using the voice of the individual they’re based on. Described as “personalized AI coaching built alongside real experts,” Portraits is built in partnership with the AI celebrity’s real-life counterpart. Our take: Google’s experiment, while it may not connect with every audience member, could prove effective at increasing time spent with Google’s AI offerings, drawing attention away from competitors like Meta AI and potentially unlocking a new revenue stream if Google monetizes in the future.
The news: OpenAI’s business user base surged 50% since February, reaching 3 million paying enterprise customers. To deepen its footprint in the space, the company released workplace features aimed squarely at Microsoft and Google, per VentureBeat. New options let employees pull and interact with cloud data from SharePoint, Dropbox, Google Drive, and more—directly in ChatGPT. Also added: Record Mode for transcribing meetings and upgrades to Codex and Deep Research. Our take: Expect ChatGPT to continue evolving from a standalone AI app to a productivity platform. Business leaders should evaluate OpenAI’s new business suite not just as a productivity upgrade, but as a strategic shift toward AI-driven business platforms.
Over 4 in 10 (44.4%) of US Adults are somewhat or very likely to use an AI tool like ChatGPT or Copilot to research potential purchases, according to April data from Attest.
Reddit is suing Anthropic for unauthorized data scraping: The case highlights growing battles over content control in the AI era.
The news: Amazon is testing humanoid delivery robots, per The Information, which could work in tandem with human drivers or as part of an autonomous fleet of delivery vehicles. The humanoid robotics team is working on incorporating large language models (LLMs) from Chinese companies DeepSeek and Alibaba so the bots can contextualize real-world surroundings. Our take: Delivery bots could help with heavy loads and ease the burden on human drivers, but Amazon might be better served with a less human form factor, such as a platform with walking legs to carry packages. The focus on humanoids could limit functionality, and bringing the uncanny valley to consumers’ front door could be off-putting.
The trend: From insight generation and content creation to media placement and regulatory reviews, generative AI (genAI) is becoming more connected to every part of pharma marketing. Our take: The tech is helping pharma marketers and ad agencies create more personalized ads and better predict ad performance—but overall, genAI usage is still pretty nascent in the industry.
The news: Medical AI startup OpenEvidence inked a multi-year agreement with JAMA Network that gives the company access to full-text content from the American Medical Association’s 13 medical journals. Our take: OpenEvidence is competing with Wolters Kluwer’s UptoDate medical information tool, which is used by a few million clinicians worldwide and has recently integrated its own AI search capabilities. One big difference between the products is that OpenEvidence is free for doctors and generates revenue through advertising. Meanwhile, UptoDate does not provide advertising opportunities. We think that OpenEvidence’s internal AI prowess could give it the leg up as long as its in-platform advertising doesn’t turn off doctors too much.
93% of business and tech leaders view it as a way to offer proactive support but caution that human empathy and oversight still matter.
Perplexity Labs signals a pivot from AI search to full-stack enterprise tools, positioning the startup as a rising competitor to Google, Microsoft, and OpenAI in workplace automation.