The AI Jungle changed in 2025. Strategy is how you win in 2026.
Back at the end of 2024, we made a set of predictions for marketing in 2025. How did we fare? Well, it turns out we nailed it. Here we look back at what we thought would happen, what did happen, and how we see the online landscape evolving over the coming year.
Our 2025 predictions
1. The Return of Voice – home assistants will make a comeback
True: You can chat to ChatGPT and Gemini in voice mode, and a host of new AI voice tools came to market in 2025.
Amazon has also released a handful of updated devices along with Alexa+, their new improved voice model. Alexa+ is classed as Early Access, but has been around since February, and more new devices launched in November. Google has been dragging their feet though, only recently announcing Gemini for Home/Nest devices. A Google Home device refresh is expected early in 2026.
Conversation is easier than ever with these devices, but the next prediction we made is the aspect that will have the biggest impact on marketing…
2. AI for businesses will mature rapidly, including agentic tools
True: If the voice assistant rollout has been slow, business tools have been supersonic.
Google stopped offering AI as a standalone plan and incorporated it into the costs of its personal and business subscriptions, recognising that AI is already inseparable from its tools. Google Ads has AI tools to scan your site for branding, content, and keywords, using that info to generate ad copy, images, and even video. Just recently, they also announced their Pomelli experiment that turns site branding into simple, curated, and editable social posts.
Tools like N8N and Jasper have brought simple AI workflows to teams and enabled individuals to operate as fast-moving jacks-of-all-trades. Existing industry stalwarts like ScreamingFrog have adapted and upgraded their tools to talk to AI APIs and combine crawled data with transformative prompts. It’s now simple to crawl and find every mention of a phrase or concept in your site’s metadata and remove, update, or refine it.
ChatGPT has been developing its business tiers and recently members of the public spotted invites to an ad platform from OpenAI being tested on iOS. They also released Atlas, their own browser, taking a shot at Google Chrome. Atlas has ChatGPT natively integrated, letting you access AI insights from just about anywhere. And more impressively, it also comes with ‘agentic’ capabilities, allowing the browser to act on your behalf, clicking links, opening new tabs, filling out forms… or buying things you need. Perplexity just launched their Comet browser, and Google are now waitlisting people for their Disco browser.
4. Google will get knocked down a peg
True! Fines galore hit the search giant: €2.95 billion from the European Commission for abuses of their ad tech dominance; €325 million from France’s data protection body for ad cookie consent abuse, and $425 million from a San Francisco jury for deceiving users about privacy controls. And Google agreed to delete all the data it has on you when you browse incognito.
But the big US vs Google (2020) case ended in the judge ruling that they would not have to break up their business by selling off the Chrome browser. That didn’t stop Perplexity offering $34.5 billion dollars for the browser. That would have covered a few of those fines…
We’re also awaiting a ruling soon for a potential break-up of their ad tech business; we wouldn’t advise holding your breath, though. Plus, Russia wants $810 million from Google (Ireland) for not giving back $1.3 billion in payments from Google (Russia) around the time the Russian government seized all of its assets.
5. New Search competitors?
True, but Goliath still stands tall. ChatGPT is now a household name, boasting over 800 million weekly users. Yet all AI platforms still only account for 0.15% of internet traffic overall. Google has a 25:1 advantage over ChatGPT in terms of queries, and Google’s traffic hasn’t been knocked at all by the appearance of so many potential competitors – perhaps because of their own AI overviews. However, click rates on Google search results are down 34.5% when an AI overview is present, showing that users are changing their behaviour.
We already know that Gen Z uses multiple sources of information to gather facts, often leading with social media before using Google. AI has just added a new platform to synthesise into that process, and other generations are doing the same. Many people now take to ChatGPT to ask longform, broad questions before going to Google with a simpler, distilled query once they’ve become more comfortably informed. In a nutshell: yes, there are new search competitors, but Google is yet to be toppled.
So, what about 2026?
If 2025 was an explosion of new AI capabilities, 2026 should, in theory, bring a focusing of that tech into real, tangible implementations. There’s also likely to be a social reckoning of just how much we’re willing to accept AI as consumers.
Creativity and strategy will become your key differentiators
As AI makes content creation easier, what will truly matter is quality, authenticity, and meaningful storytelling. This is where we’re heading as an agency: digging deep into a business’s why, their true USPs, and crafting a compelling story to engage new audiences. It’s a two-fold approach: crafting the brand and messaging first, then delivering the right content through the right channels.
A new age of SEO and AIO for discoverability
Businesses that use AI to understand customer behaviour, create high-quality human-centric content, and adapt quickly to data insights will gain greater visibility, while those relying on outdated tactics will see their reach decline.
AI-driven ranking systems will become even more personalised, predictive, and intent-focused. Instead of rewarding volume or frequency, algorithms will prioritise genuine expertise, real customer engagement, and trust signals across multiple platforms. The competitive advantage will shift from ‘who posts the most’ to ‘who delivers the most relevant and valuable experience’.
New laws to help consumers tell real content and AI content apart
Think GDPR or cookie laws, multiplied by 100. Governments will begin making efforts to mandate that AI-generated assets are watermarked or otherwise identifiable in the face of AI video and voice content being used at scale. New models like OpenAI’s Sora 2 and Google’s Veo 3 are uncannily realistic at times and AI is already being used to commit crime. The technology to limit AI misuse is in its infancy, so legislative intervention is long overdue.
Advertising will move from sponsored sections to dynamic inserts in conversations with AI
Right now, AI can help you write an ad and generate lifestyle imagery from your product shots. ChatGPT is testing an ads platform on iOS, and Google is even rumoured to be testing fully generated ads. But where are all these ads going to go?
On Search, ads can easily be camouflaged among other results, encouraging users to at least consider them. In conversation with AI, these old ad formats would (and do) stick out like a sore thumb. And when something is so abrasive, users quickly learn to tune it out, scrolling and skipping to the part of the user interface they want to see. So expect ads to be inside AI responses, seamlessly added in as [sponsored] hyperlinks to products being mentioned or recommended.
The problem will be how AI tools incorporate their own model’s weighting with the intrusion of forced weightings from ad auctions. If ChatGPT wants to recommend three products, and an advertiser not shortlisted for those podium positions stumps up the ad budget, will ChatGPT then recommend a bad product as a sponsored link? Only time will tell.
Companies will invade public life in new ways to train models
We’re dealing with a massive oversaturation of new ways to interact with the web – and with information more generally. Not everything is proving to be useful to end users. AI ‘slop’ is everywhere (word of the year!), and anyone can spin up a half-baked idea in minutes.
Even companies like Meta are guilty of this, forcing their way into our lives with their Ray-Ban partnership to bring augmented reality glasses to the public domain. The offer is a handy assistant on your head that looks ‘cool’ too. But the reality is a dangerous invasion of everyone else’s privacy (for which Meta offloads responsibility onto you), and a tool for Meta to gather video training data for their AI projects.
Make no mistake: companies are vying to commodify our lives in ever more novel ways. The first in-home humanoid robot is launching in 2026, and the creators are upfront about early adopters being a vessel for their device to gather the training data it needs as it goes. The reality is that a remote operator will be controlling it and seeing into your home for 90%+ of its complex tasks until it has enough data. The data? Human activity. Washing up, unpacking shopping, tidying, feeding your pets, and so on – in 2026 corporations will be watching your every move.
What do you think? Let us know your hopes, fears and predictions for 2026, and if you’re looking for that creative and strategic edge to help your business ROAR louder than your competitors, contact us for a chat.