Around the World – The AI Issue

There’s no denying AI’s effect across media and marketing is not just topical but pervasive. This week, media expert Antony Young uses his regular column to go all in with consequential AI stories impacting our industry right now.

This week:

• Tilly Norwood: Thief or re-inventor?
• OpenAI’s new app a challenge to TikTok
• Meta plans to use AI data to improve ad targeting
• The changing world of SEO: here comes AEO and GEO
• Agency holding companies are shifting away from an ‘agent’ model
• The Economist rolls out a series of initiatives to AI-proof its business

Is Tilly Norwood stealing from actors or re-inventing storytelling?

The backlash against Tilly Norwood, the AI-generated “actor,” has drawn sharp criticism from SAG-AFTRA, UK Actors Union Equity, and stars like Emily Blunt who argue that synthetic creations threaten the integrity and livelihood of human performers.

Here’s an interesting and opposing take by media writer Shelly Palmer who argues they are missing the point. Tilly isn’t an actor in any traditional sense. That it’s a statistical construct trained on vast human data, not individual performances. Rather than stealing from actors, AI like Tilly mimics aggregate human behaviour, challenging the very definition of acting and storytelling.

For marketers and advertising agencies, this signals a seismic shift: synthetic personas can now emulate emotional nuance and cultural relevance. They offer scalable, cost-effective content creation. The question Palmer raises isn’t whether AI will replace actors, but how we will need to redefine human value and creativity in a media world that will be shaped by algorithmic storytelling.

OpenAI’s video app Sora to challenge TikTok

OpenAI’s new app Sora has ignited both excitement and concern by blending AI-generated video creation with a TikTok-style social media experience. Users can generate hyper-realistic short-form clips featuring themselves or others often with surreal or humorous scenarios with its “Cameos” feature, which allows facial likeness uploads. A sidebar on the right lets users engage by liking, commenting, or remixing videos.

It appears OpenAI is betting that Sora 2 will redefine how people engage with AI video, much like ChatGPT did for AI text. Concerns over deepfakes and copyright misuse have been being raised by the Motion Picture Association demanding Open AI provide safeguards for creators copyrights that are being used without consent.  

Meta plans to use your AI data to improve ad targeting

It’s coming! Meta will begin using data from user interactions with its AI products including chatbots, smart glasses, and image generators to fuel targeted advertising across Facebook and Instagram. They plan to update their privacy policy globally by December 16 to allow them to access this user data. 

The new policy applies globally, except for users in South Korea, the United Kingdom, and the European Union, where privacy laws prevent this type of data collection. With over a billion monthly AI chats, Meta sees AI data as a new signal to refine ad targeting, which users cannot opt out. While sensitive topics such as religion and politics are excluded from ad personalisation, the move underscores how free AI tools are going to be increasingly monetised. Google and OpenAI will no doubt follow.   

Moving from SEO to AEO and GEO

Sorry about this acronym-fest! As AI transforms how people search for and consume information, SEO is evolving from keywords to a more sophisticated, intent-driven discipline. Traditional tactics like keyword stuffing are becoming obsolete, with AI-powered search engines now prioritising content that demonstrates clarity, depth, and genuine usefulness. Features like Google’s AI Overviews are reducing click-through rates by surfacing instant answers, making it critical for brands to offer unique value beyond what AI can summarise.

Experts say to stay discoverable, marketers must adopt hybrid strategies that blend classic SEO with Answer Engine Optimisation (AEO) and Generative Engine Optimisation (GEO), regularly updated content which AI engines love, and design content for platforms like YouTube, TikTok, and AI assistants. In this new landscape, experts argue success hinges on earning trust and authority rather than gaming algorithms.

Agency holding companies shifting away from an ‘agent’ model 

Forrester last week forecast that 2026 will mark a dramatic transformation for agencies, with 15% of jobs lost and a shift away from the traditional “agent” model. Instead, agencies are evolving into product owners, media resellers, and tech developers, driven by pressures from shrinking margins, in-housing, procurement demands, and the rise of AI. 

Forrester are predicting the end of agencies as independent client representatives. Several industry developments are taking place that are accelerating that change. This includes further mergers and consolidation of global agency groups such as Omnicom and IPG; creative agencies taking control of influencers and content creators; and the mainstreaming of principal media, i.e. media agencies taking positions on media space and on-selling to clients as principals instead of agents. 

Critics argue these developments spearheaded by the marketing holding companies are designed to enhance profitability but conflict with representing the clients‘ best interest.  

How The Economist plans to AI-proof its business

To navigate the risks of AI disintermediation The Economist is rolling out a series of initiatives designed to deepen audience engagement and protect its content. One major move is the launch of Insider, a twice-weekly video series featuring its senior editors. These 30–60 minute episodes, free to subscribers, offer direct access to the newsroom’s thinking and are designed to be difficult for AI to replicate, while opening new inventory for advertisers.

The publisher is also developing its own proprietary AI tools. One prototype is a ring-fenced chatbot built into its website, allowing subscribers to query articles and receive nuanced responses grounded in its archive and editorial expertise. This aims to create a more interactive subscriber experiences whether through five-minute summaries, deep dives, or cross-format content. 

To protect its journalism, The Economist is actively blocking AI bots from scraping its content via cybersecurity protection platform Cloudflare and refusing to license its material to large language models. Instead, it’s exploring controlled visibility on AI platforms for brand awareness, not traffic or monetisation. It’s also experimenting with niche Substack newsletters to test demand for new topic areas without cannibalising its core subscription.

The post Around the World – The AI Issue appeared first on stoppress.co.nz.

stoppress.co.nz