KIMP’s Picks September 25th: Weekly Highlights In Design, Marketing, AI, & More

Welcome to KIMP’s Picks’ September 25th edition. We are exploring what’s buzzing on social media – the big AI news that people are talking about, changes coming to popular social media platforms and more.  

From OpenAI’s new GPT-6 model additions to YouTube’s latest suite of AI discovery and advertising tools, we’ve gathered the top news shaping creative strategies. Read on to discover how these shifts will impact your workflow and campaigns this week.

In the Marketing Realm

YouTube Rethink 2026 expands AI, commerce and measurement tools for advertisers 

YouTube has outlined several advertising updates for 2026, covering AI-powered shopping, Demand Gen, product feeds, creator insights, messaging, and brand measurement.

Business Agent for YouTube Ads is entering open beta in the US. The AI assistant appears below an ad, allowing viewers to ask product questions using information from an advertiser’s product catalog and website. 

YouTube is also testing clickable disclaimer overlays for Demand Gen campaigns, giving advertisers in sensitive sectors such as finance, healthcare, and pharma a way to provide required disclosures. The company is piloting click-to-message ads in Southeast Asia, Brazil, and India. This allows users to start conversations with businesses from video ads. Product feed updates add Things to Do feeds and travel-intent personalization. 

What a brand mascot can do amidst AI slop 

I Can’t Believe It’s Not Butter is using its new mascot, Tubz, to give the brand’s latest chapter a more personal angle. Tubz debuted in April and now appears as the fictional author of Lifting the Lid, a 78-page self-help book available on Amazon. Social media producer Cassy White and a ghostwriter created the book, which includes a foreword from former brand ambassador Fabio.

The campaign draws on the brand’s history. I Can’t Believe It’s Not Butter launched in 1981 with a lower-fat alternative to butter and became a major grocery brand. Its momentum later declined as butter regained popularity.

The brand is now part of Flora Food Group and recently underwent a makeover and reformulation. The updated spread contains vitamins A, D, E and B12, plus omega-3 ALA. Its marketing also highlights its potential role for consumers taking GLP-1 medications.

Tubz’s story mirrors that comeback narrative, using humor, nostalgia and personal growth to reconnect with consumers. The campaign shows how mascots, nostalgia and product changes can work together to build a fresh narrative around an established brand.

SEO tasks worth automating right now 

SEO automation is becoming more accessible as AI tools and traditional workflow platforms can now handle a wider range of repetitive tasks. Search teams can combine tools such as Zapier and Make with LLMs to automate parts of technical SEO, keyword research, content workflows, link management, competitor monitoring, reporting and lead qualification.

A recent article from Search Engine Land highlights practical uses including automated technical checks, structured data validation, keyword intent and lifecycle tracking, topical mapping, content briefs, meta tags, image alt text, internal link suggestions and SEO testing. AI can also help identify entity gaps and review generated content against predefined quality rules.

The approach still requires human oversight. LLMs can produce inconsistent or incorrect outputs, so the article recommends using AI for tasks where results can be checked easily. 

Retail media is evolving and brands need to adapt to the changes 

Retail media is expanding beyond ads placed directly on retailer websites and apps. As consumers move between search, social, streaming, delivery platforms, marketplaces and loyalty programs, brands are using first-party transaction data to reach shoppers across more stages of the buying journey.

A recent MarTech article discusses how this shift is putting more focus on purchase intent. Audience size still matters, but marketers are also assessing how close an audience is to making a purchase, along with buying frequency, category depth and data quality. 

Measurement is changing as well. ROAS remains a common benchmark, while marketers are looking for stronger evidence of incremental sales and business growth. The article highlights incrementality testing, closed-loop attribution, customer lifetime value and cross-channel measurement as tools for evaluating impact.

The growth of commerce media creates more opportunities to combine purchase data, high-intent audiences and stronger measurement across the customer journey.

Google expands AI Max with support for more languages and new Search reporting 

Google is adding new controls and reporting features to AI Max campaigns in Google Ads. The closed beta of AI Brief is expanding to Dutch, French, German, Italian, Japanese, Portuguese and Spanish. The feature lets advertisers provide AI Max with detailed guidance about their business, target audience and key messaging using natural language.

Google is also developing a unified reporting view for the Search ads journey. The report will connect the main steps of an ad interaction, showing the search terms that triggered ads, the creative assets shown to users and the website destination they reached.

The new reporting is designed to give advertisers more visibility into how AI-driven campaigns perform and how users move from search to a website. 

In the AI Realm

OpenAI expands the GPT-6 family with new affordable models 

OpenAI has expanded its GPT-6 model family with GPT-6 Sol and GPT-6 Luna. These models are built to be cost-effective additions alongside GPT-6 Astra. According to OpenAI, both models use similar training methods to Astra, with improvements across professional work, factuality, coding, computer use and alignment.

API pricing for Sol and Luna is being cut by 50% compared with the promotional pricing of their GPT-5.6 counterparts.

With updated caching capabilities, these models reportedly also show stronger performance on several professional, coding and computer-use benchmarks, alongside improvements in factual reliability. 

Gemini expands ecosystem with new connected apps 

Google is rolling out a new wave of integrations for the Gemini app, bringing popular third-party tools directly into your workspace.

Users can now manage workflows, design creative assets, and handle lifestyle tasks without switching tabs. Productivity favorites like Airtable, Linear, and monday.com are joined by creative giants like Adobe and Squarespace, alongside lifestyle platforms such as Peloton and Apartments.com.

To use them, simply head to your settings to connect your tools, or type an @ mention directly in your chat to pull them into the conversation.

This update transforms AI assistants into central operating hubs, creating new opportunities for software platforms to embed themselves seamlessly into daily consumer and professional workflows.

Google unveils Gemini 3.8 Flash TTS for advanced voice generation 

Google has launched the Gemini 3.8 Flash TTS and Flash-Lite TTS models, bringing hyper-realistic and expressive audio generation to its ecosystem.

The new models let creators design custom voices from scratch using simple text prompts or even with quick 30-second audio samples. Users have control over pacing, accents, emotional nuances, and multi-speaker scene staging across more than 100 languages.

Safety features include mandatory consent verification for voice replication and SynthID watermarking to ensure transparency and prevent misuse. Developers can access these tools via Google AI Studio and the Gemini API.

Now all Google users can access AI video generation in Vids 

Google is expanding access to Google Vids powered by the Gemini Omni 1.1 model, allowing anyone with a Google or Google Workspace account to create AI-generated videos at no cost. The update aims to make video production more accessible for businesses, creators, organizations and everyday users without requiring advanced editing skills.

New controls give users more flexibility over the creative process. Videos can be generated in 1080p HD, existing AI clips can be upscaled, and scenes can be extended while maintaining visual consistency. Users can also set precise clip durations to better match storytelling and voiceovers.

To support transparency, every AI-generated clip includes a SynthID digital watermark embedded in the video. 

Stronger agentic coding capabilities at a lower cost – Claude Opus 5.5

Anthropic has released Claude Opus 5.5, the first model in its new Claude 5.5 family. The model reportedly delivers performance comparable to Claude Fable 5.1 on many tasks while reducing operating costs by 40% compared with Opus 5.

The update focuses on agentic coding, computer use, knowledge work, efficiency and safety. Anthropic reports that Opus 5.5 performs strongly on software engineering tasks, including large-scale code migrations, code audits and performance optimization. The company also highlights improvements in business analysis, research workflows and long-form knowledge tasks.

Meta announces Muse realtime avatar 

Meta has introduced Muse Realtime Avatar, a new AI system that adds expressive visual avatars to its Muse Realtime Voice platform. The technology can animate a wide range of characters, including photographic portraits, illustrations, animals and objects, enabling them to participate in live conversations with synchronized speech, facial expressions, gestures and body movement.

The system combines conversational AI, voice generation and avatar rendering through a shared stream of speech tokens. This approach helps keep voice, lip sync and visual expressions aligned while maintaining consistent appearance and behavior throughout extended interactions.

ElevenLabs Studio 4.0 simplifies end-to-end AI video generation 

ElevenLabs has launched Studio 4.0, a major update to its AI-native video editor that combines content generation and editing in a single workflow. Users can now create video, images, voiceovers, music and sound effects directly inside a project, eliminating the need to move between multiple production tools. 

The generated assets are placed directly on the editing timeline, so it’s easier to refine scenes without exporting or rebuilding projects. The updated editor also introduces frame-level zoom, clip snapping and timeline-based caption editing, allowing creators to adjust subtitles and video elements with greater precision.

Social Media Feature Updates and Algorithm Changes

YouTube announces a host of new AI tools for creators 

YouTube is expanding YouTube Studio with AI-powered tools covering research, video feedback, thumbnails, editing and community management.

For instance, the new Insights and Research features will help creators analyze past performance and identify outlier videos for future ideas. And YouTube Studio can also review drafts and provide feedback on pacing, structure and storytelling. 

Additionally, creators will be able to generate channel-matched thumbnails, while Dynamic Thumbnails can show different options to audience segments. YouTube is also testing video A/B testing for up to three different cuts to compare hooks.

Another notable update is that Gemini is being integrated into a conversational editing assistant for Shorts and the YouTube Create app. So, creators can use prompts to build drafts, reorder frames, trim clips and sync music before switching back to manual editing.

YouTube launches Custom Feeds and interactive shopping tools 

YouTube is rolling out major updates to enhance how users find and interact with content on the platform. A new Custom Feeds feature allows viewers to build dedicated homepage tabs based on specific moods, routines, or personal interests. 

Additionally, YouTube is upgrading the Ask YouTube tool to serve as an interactive shopping guide. When searching for product reviews, users will receive organized video recommendations and detailed comparison tables. From there, viewers can also continue asking follow-up questions directly on the watch page while exploring creator recommendations.  

Conversational AI discovery in YouTube Music 

YouTube Music is adding new AI-powered discovery features designed to help users find artists, songs and podcasts across the platform. According to YouTube, users discovered a new-to-them artist more than 450 million times in July 2026.

Also, Ask Music is expanding inside the YouTube Music app with conversational controls across more than 300 million songs and podcasts. Moreover, users can customize their queue and explore additional information about artists, giving them more ways to interact with the platform’s music catalog.

YouTube Music and Premium subscribers will also get Your Podcast Lineup, a personalized weekly audio guide that previews recommended shows and explains why they may appeal to each listener. The experience is designed to work hands-free, with users able to listen without looking at the screen.

X announces X Numbers for direct messaging 

X has rolled out X Numbers, an optional private code that lets people message or call an account through Chat even when they do not follow each other. The feature gives users another way to control who can reach them without opening their inbox to general message requests.

When enabled, an X Number lets someone contact an account directly in the inbox, bypassing message requests. Calling is also supported, depending on the account’s call settings. Users can hide their number from others, turn it off, or refresh it to create a new code and invalidate the previous one.

X Numbers are not phone numbers and are separate from Chat safety numbers. Users can find the feature through Chat’s inbox dropdown and share the code selectively.

X revamps Livestream API for enhanced broadcast controls 

X has overhauled its Livestream API to give authorized third-party applications more control over live broadcasts. The API now covers the broadcast lifecycle, from creating and publishing a stream to monitoring and ending it.

Additionally, developers can manage persistent RTMP and RTMPS stream sources that can be reused across multiple broadcasts. The platform also supports scheduled broadcasts, including one-time and recurring events.

Live chat is another major part of the update. Applications can read chat messages through the X Activity API, send messages to live broadcasts, and moderate conversations by removing messages or muting users.

Wrapping Up

From cost-effective AI models to interactive social commerce, this week clearly showed that the industry is leaning into smarter, integrated experiences. Leveraging these emerging tools can help streamline your daily operations and sharpen your marketing strategy. Try integrating one of these highlights into your workflow this coming week. See you next week with the next KIMP’s Picks edition. Until next Friday, keep creating and innovating!