AI has continued to dominate the headlines, with rapid advances and fresh launches over the past weeks. One thing we haven’t seen much of, though — even as we head into Christmas and the big brand films land — is GenAI being used front and centre in festive campaigns. Aside from a couple of notable exceptions like Coca-Cola’s “Holidays Are Coming” and McDonald’s It’s the Most Terrible Time of the Year, most seasonal creative still looks very traditionally produced. 

Where things are moving quickly is inside the platforms themselves, with a wave of product updates and improvements from major media players. Read on for the latest.  

TikTok Empowers Users with New AI Controls and Creative Tools 

TikTok is adding an AI slider that lets people choose how much AI-generated content appears in their feed, alongside clearer labelling, watermarking, and digital-literacy efforts. It’s also using C2PA Content Credentials to embed metadata that shows when content is AI-generated, and says these steps have helped label over 1.3 billion videos so far. TikTok has also upgraded its creator tools with Smart Split, which turns longer videos into multiple TikTok-ready clips with reframing, captions and transcripts, and AI Outline, which generates a customisable six-part structure plus titles, hooks and scripts from prompts. 

Why it matters 

As synthetic media floods social feeds, control and provenance become essential. Giving users an easy way to reduce AI content if they want, and making AI-generated posts easier to spot, helps protect trust for audiences and brand safety for advertisers. At the same time, TikTok is adding more AI features to support creators (building on tools like TikTok Symphony and TikTok Alive), while pushing AI creation forward with clearer guardrails. 

Google’s Gemini 3 Sets New Standard in AI Performance and Competition 

Google launched Gemini 3 and, for now, it looks like it’s taken the lead in the AI race. Early coverage points to a real step forward in reasoning, multimodal performance and more agent-like capability, shifting attention away from OpenAI’s ChatGPT. Gemini 3 is also sitting at or near the top of LMArena’s leaderboard (one of the most watched public signals of which model feels best in head-to-head use) and it helps explain why reports suggest OpenAI is treating this as genuine competitive pressure, with the Financial Times and The Guardian, among others,  describing an internal ‘code red’ push to sharpen ChatGPT. 

Why it matters: 

Gemini 3 could be the burst of publicity Google needs to push Gemini into the mainstream and grow its share worldwide as a single, default app (versus ChatGPT) — turning it from a good model into a habit. But in professional GenAI work, it’s rarely about picking one winner. Most teams get the best results by using a mix of models and comparing outputs, because each has different strengths. And with models improving all the time, today’s lead can be tomorrow’s draw. That said, Gemini 3 should definitely be on the table for teams to trial and roll into their toolset where it performs best. 

Pinterest Launches AI Shopping Assistant  

Pinterest is rolling out an AI shopping assistant that lets users chat (voice-first, at launch) to get personalised outfit ideas and product suggestions, pulling context from what they’ve saved and the pins they’re looking at 

Why it matters: 

This is another sign that shopping is shifting from search-and-scroll to conversational discovery,  where people describe the vibe they want and the assistant does the filtering. Pinterest has an edge here because it sits on strong intent signals (boards, saves, collages) and can turn them into more relevant recommendations without users having to start from scratch. If it works, it keeps users in-app longer and moves Pinterest further from ‘inspiration’ into an end-to-end shopping journey. 

Google and OpenAI Bring Conversational Shopping to Search  

Google and OpenAI are pushing shopping deeper into the chat interface. Google is rolling out more conversational shopping inside Search’s AI Mode to help people refine what they want through back-and-forth prompts, while ChatGPT now includes ‘shopping research’, which asks clarifying questions, searches the web using higher-quality sources (and can use your chat history), then produces a tailored buyer’s guide. It’s available on mobile and web across plan tiers, with ‘nearly unlimited’ usage during the holiday period. 

Why it matters: 

This shifts product discovery from keywords and filters to conversation, and raises the bar for brands: product information, reviews and comparisons need to be clear, consistent and easy for AI systems to interpret and trust. It also increases the need for transparency around what sources are used and how recommendations are made, because these tools can influence purchasing decisions quickly and at scale. 

Getty Images and Perplexity Sign AI Image Partnership 

Getty Images and Perplexity have signed a global, multi-year licensing agreement that allows Perplexity to display Getty’s editorial and creative images across its AI-powered search and discovery tools, with an emphasis on clear attribution (credits and source links).  

Why it matters:  

it’s a clear signal that the industry is moving away from ‘use it because it’s online’ scraping towards structured, paid licensing for high-value content in AI products. For AI search, that improves trust (users can see where visuals come from) and reduces legal risk; for publishers and rights-holders, it’s a route to monetisation and control. It also reinforces a broader trend of AI firms signing content deals as copyright scrutiny and litigation rises.  

Rightmove Rolls Out AI Tools to Transform Home Search and Seller Insights 

Rightmove has published a round of new AI-powered tools, reflecting its broader strategy to embed AI across its services, which most of them rolling out gradually  

  • Vendor Prediction Model: Helps agents spot likely future sellers earlier, using Rightmove’s data, and is integrated into Opportunity Manager and Rightmove Discover. 
  • AI Keywords: Lets movers search in natural language (e.g., “river view”, “exposed brick”) rather than relying only on standard filters.  
  • Style with AI: Lets users visualise changes to a home’s look in listing images (such as décor and lighting), including removing furniture. 
  • Renovation Cost Estimator: Gives movers upfront estimated renovation costs to support budgeting and more informed decisions before booking viewings, and helps highlight the potential of properties that need updating. 

Why it matters: 

AI is moving quickly, so building it directly into products,  and keeping more of the customer journey on-platform, is becoming essential. As experiences shift towards conversational interfaces, chat-like, natural language search should better match how people actually browse, so it’s important businesses design with that in mind.  

At the same time, tools that bring buying decisions into the same website can only improve the overall experience, keeping consumers better informed and on your site for longer, rather than them going elsewhere to find key information.  

And on the business side, building predictive models are increasingly a must for more cost-effective planning and targeting,  but they only work with clean, joined-up data and a solid reporting layer that can scale