Catch up on select AI news and developments from the past week or so (in no particular order):
Madison Avenue braces for AI disruption as automation reshapes ad industry. At Cannes Lions 2025, advertising leaders have confronted the accelerating impact of AI on their business, from job cuts to agency mergers. Executives at WPP, Omnicom, and Meta see automation transforming both ad creation and media buying, while Google’s Veo 3 and Meta’s generative tools enable brands to rapidly scale content. AI may shrink traditional agency roles and deepen reliance on tech giants. Still, leaders argue that human creativity remains vital. Importance for marketers: As AI remakes advertising workflows and media planning, marketers must prepare for automation’s growing role while doubling-down on creativity and strategic oversight.
Google and others drop Scale AI after Meta deal raises conflict concerns. Google, OpenAI, Microsoft, and xAI are cutting ties with Scale AI following Meta’s 49% stake in the data-labeling firm. Concerns over competitive exposure have prompted these companies to seek neutral vendors or in-house teams for sensitive AI model training. Scale AI’s CEO is joining Meta, intensifying industry fears of data leaks. Scale may now pivot to its applications business for survival. Importance for marketers: With data privacy and competitive intelligence at stake, this shift underscores the rising need for neutral, secure data vendors in AI development workflows.
Google AI Mode rolls out to all US users, reshaping search experiences. Google’s AI Mode, first revealed at I/O 2025, is now live for US users, offering a multimodal, AI-powered search interface with a “query fan-out” system for deeper exploration. It supports conversational follow-ups and could significantly shift user behavior. Search Console data will now lump AI Mode activity into standard web search metrics. Importance for marketers: Marketers should prepare for traffic shifts and evolving search behavior, as AI Mode changes how users discover and engage with content.
OpenAI considers antitrust complaint against Microsoft, its biggest backer. OpenAI may file an antitrust complaint against Microsoft over cloud exclusivity and investment terms. Tensions have grown amid OpenAI’s shift to a public benefit corporation and its desire to renegotiate Microsoft’s rights to host its models. The case could disrupt one of the most important AI partnerships. Importance for marketers: A fallout could affect the availability and cost of OpenAI’s tools within Azure, impacting marketing platforms and integrations relying on that ecosystem.
Big Tech’s AI ad tools threaten the traditional agency model. Meta, Google, and Amazon are building AI tools to automate ad creation and media buying, bypassing agencies. Brands like LVMH are already adopting in-house AI capabilities. Agencies are responding with M&A activity to stay relevant. Importance for marketers: Agencies must evolve fast or risk obsolescence; brands have new opportunities to take ownership of campaign execution with precision and efficiency.
Midjourney debuts V1 AI video generation model amid legal pressures. Midjourney launched V1, its first video generation model, allowing users to turn images into 5–21 second video clips. The launch follows lawsuits from Disney and Universal over IP use. V1 is only available via Discord and targets creatives over commercial users. Importance for marketers: The tool expands creative options for short-form video content, but IP concerns remain a serious risk in brand-safe environments.
Salesforce study finds LLM agents weak in CRM tasks and confidentiality. Salesforce researchers report that LLM agents score just 58% on single-step CRM tasks and drop to 35% on multi-step ones. They also perform poorly with handling confidential data. Importance for marketers: This highlights the gap between hype and capability for AI agents in CRM, signaling caution for marketers integrating LLMs into customer interactions.
Reddit rolls out AI ad tools to tap into community insights. Reddit introduced “Reddit Insights powered by Community Intelligence” and “Conversation Summary Add-ons” to help brands align ads with community sentiment. Publicis is an early adopter, with wider rollout expected in July. Importance for marketers: These tools offer a new way to surface authentic user engagement and improve contextual ad relevance.
Google rolls out Gemini 2.5 Pro, Flash, and Flash-Lite to all users. Gemini 2.5 Pro and Flash are now stable and widely available, with Flash-Lite joining in preview. These reasoning models support adjustable “thinking budgets” and have been integrated into Google tools. Importance for marketers: Access to more capable reasoning models could enhance automation in content creation, personalization, and marketing analysis.
OpenAI confirms departure from Scale AI over neutrality concerns. OpenAI has cut ties with Scale AI, joining Google and Microsoft in moving to neutral or in-house data solutions after Meta’s stake. Scale aims to shift toward applications and government work. Importance for marketers: Reinforces the critical importance of data governance and neutrality when training proprietary AI models.
Adobe launches mobile Firefly app with third-party model support. Adobe released its Firefly AI image-generation app for iOS and Android, integrating its own models with partners like Ideogram and Runway. The app allows unlimited basic use, with premium options available. Importance for marketers: Marketers can now generate creative assets on the go with commercially safe models, helping speed up campaign production workflows.
Gemini app adds video upload and analysis across platforms. Google has rolled out video upload and analysis features in its Gemini app for both free and paid users. Users can ask questions about uploaded clips. Importance for marketers: This feature unlocks new creative and analytical possibilities for video-based marketing insights and repurposing video content in campaigns.
Watchdog challenges Microsoft’s Copilot productivity claims. The NAD has called on Microsoft to clarify or retract productivity claims in Copilot ads, citing lack of substantiation and confusing branding. Microsoft will comply. Importance for marketers: Messaging about AI productivity must be grounded in verifiable data—marketers should audit claims to avoid reputational risk.
Apple Intelligence nears China launch via Alibaba’s Qwen3 model. Apple’s AI features are nearing release in China, thanks to a new version of Alibaba’s Qwen3 model made compatible with Apple platforms. The move aligns with local regulations. Importance for marketers: A China launch could boost iPhone sales and introduce new AI features for marketers targeting Chinese consumers.
Oracle adds xAI’s Grok 3 model to cloud lineup for businesses. Oracle Cloud will offer Grok 3 to enterprise customers seeking AI tools under strict data governance. It joins models from Meta, Mistral, and Cohere in Oracle’s catalog. Importance for marketers: The expansion of AI model options in secure cloud environments makes it easier for marketers to integrate advanced tools while complying with internal policies.
Nvidia’s sovereign AI push gains traction with EU leaders. Nvidia’s Jensen Huang promoted sovereign AI during a European tour, securing support from UK, France, and Germany. Mistral and Nvidia plan joint data centers to reduce US dependency. Importance for marketers: Expect more regional AI infrastructure and tools optimized for local regulations and languages, reshaping how global marketers deploy AI.
Mastodon updates rules to ban AI model training on user data. Mastodon now explicitly prohibits scraping for AI model training on its .social instance. The change reflects growing concerns about unauthorized data use. Importance for marketers: As platforms restrict AI training access, data sourcing will become more complex—marketers must verify how AI tools are trained to avoid ethical or legal issues.
Pope Leo XIV warns AI could harm youth development. The Vatican voiced concern over AI’s impact on children’s neurological and spiritual development, calling for ethical AI use that protects human dignity. Importance for marketers: Reinforces growing global scrutiny over AI and its societal influence—brands should consider these concerns when promoting AI-powered products or features.
You can find the previous issue of AI Update here.
Editor’s
note: GPT-4o was used to help compile this week’s AI
Update.
