Speed kills (the competition)🏃‍♂️

Hello marketers. Welcome to AI Marketing School, where we dish out the latest and greatest in AI-powered marketing.

In this week’s issue:

• AI Marketing Update: Cannes Lions thrusts AI into the spotlight

• The Stack: Get lost in deep research

• AI Events: Our recommended AI marketing events for networking and connections

Onwards!

AI MARKETING UPDATE

Controversy in Cannes

Every June, the global advertising elite descends on the French Riviera for the Cannes Lions, the industry's most prestigious awards festival. Think of it as the Oscars of advertising.

The Grand Prix is the ultimate prize – there's only one per category, and winning one can transform an agency overnight. It's not just a trophy; it's a business-generating machine.

This year's festival was supposed to celebrate creativity in the age of AI. Instead, it became the epicentre of the industry's biggest authenticity crisis in decades.

The nuclear story was that of Brazilian agency DM9's Creative Data Grand Prix being stripped after they used AI to fabricate CNN Brasil news coverage, triggering resignations and forcing Cannes to completely overhaul its AI rules. 

The AI manipulation that shocked Cannes

Cannes Lions confirmed that DM9's Creative Data Grand Prix winner, 'Efficient Way to Pay' for Consul, was withdrawn after discovering that AI-generated and manipulated content was used within the case film to simulate real-world events and campaign outcomes.

Specifically, it faked footage from a CNN Brasil broadcast.

Here’s what happened:

• DM9 used AI to create fake CNN Brasil news footage

• Made it look like their campaign got major media coverage

• CNN Brasil issued a formal complaint against the agency

• The jury was completely fooled during judging

• 12 total awards were withdrawn across three DM9 campaigns

• The Co-president and chief creative officer, Ícaro Doria, stepped down, taking "full responsibility for the incident.

You can view the excerpts below, which were added to a 2018 TED Talk by United States Senator Deandrea Salvador. It also features a modified headline from CNN Brasil, with changes to the presentation of then-anchor Gloria Vanique.

The judging process was supposed to be rigorous – 400 initial entries whittled down to 80 shortlisted submissions over "60 hours of deep, focused work over weekends."

However, it fooled everyone. Jury member Ndeye Diagne said, “It was a unanimous vote for DM9... everyone thought the idea behind 'Efficient Way to Pay' was so amazing and beautiful. It was a Grand Prix for us in terms of the idea”…”it was "impossible to see, you couldn't imagine that people would do something like that."

In the fallout, Cannes completely rewrote the playbook with enhanced measures, including:

• Enhanced Code of Conduct that must be signed by all participating organizations

• Mandatory AI disclosure as part of the entry process, with non-disclosure constituting grounds for disqualification

• Content detection tools to help identify manipulated case films and materials

• Dedicated review committee comprised of experts in AI, ethics, and content integrity

Rather than the use of AI itself, the use of AI to deceive was the problem here.

Is the outrage justified? Probably. Imagine normalizing advertising that outright lies to its viewers using fake news coverage…you can’t go there, not ethically or legally.

Cannes has had many controversies over the years, so you’d be forgiven that the whole thing is a bit of a pretentious joke at this stage.

However, there are some important insights we can garner from the whole thing:

1. AI vs humanity: A central tension

From the festival's key sessions, one theme dominated: "AI needs humans," but nobody could specify what that actually meant. 

As Nick Law from Accenture Song observed: "A handwavy statement that is true, but achingly unspecific. The details of what we make, who makes it and how we make it, are still fuzzy."

2. The three worlds problems

Law identified a concerning split: "I can't remember a Cannes being so divided into three worlds - the tech platforms, the agencies and the brands. The platforms were bubbling with optimism as they demonstrated more concrete creative applications of genAI.”

“The agencies felt lost, unable to articulate a coherent model for a future that is arriving too fast. And the brands seemed to be watching and waiting with a detached curiosity."

Makes sense. Platforms (e.g. the gen AI vendors) are fervently pushing their technologies. In many senses, their business models benefit from the demise of agencies, who find themselves in a precarious position.

Brands may struggle to leverage the best of gen AI though — agencies need to show them the way. That is one way they can definitely survive.

3. AI video generation breakthrough

YouTube CEO Neal Mohan made the biggest AI announcement of the festival: Veo 3 will be coming to YouTube Shorts this summer.

The Google DeepMind video generation model "vastly improves video quality and incorporates audio" - representing a massive leap in AI video creation. 

Many left knowing that soon, videos will be created without even needing a camera. 

4. Meta's 11 new AI ad tools

Meta unveiled its most comprehensive AI advertising suite yet, including:

• Brand-consistent automation with logos, fonts, and color palettes

• Image-to-video generation that stitches up to 20 product photos into polished multi-scene clips

• Business AI chatbots for Instagram Stories and Facebook Reels with voice interaction

• Creative Sticker CTAs and AI-powered Virtual Try-On features

Advantage+ sales campaigns already boost return on ad spend by an average of 22%, while backend AI innovations improve ad conversions by up to 5%.

5. Adobe's customer experience 

Adobe launched Customer Experience Orchestration (CXO), fusing creativity, marketing and AI to deliver intelligent, scalable customer experiences. Key innovations include:

• GenStudio for Performance Marketing: AI-first application for creating on-brand campaign content

• Firefly Video Model: Commercially safe short-form video creation

• Firefly Services APIs: Streamlining color grading and resizing workflows

Firefly marketing itself as the commercially safe video tool is a key announcement.

When video generation tools go mainstream, it will be interesting to see if and how tech companies avoid copyright issues. 

Creators could absorb some of the legal liabilities if people naively or accidentally use AI-generated copyrighted footage (e.g., a clip featuring Marvel characters or Mickey Mouse) to promote their brand. “I generated it with AI!” will not always be a suitable legal defence.

6. Microsoft's Brand Agents

Microsoft introduced Brand Agents – AI-powered digital assistants that act as personalized extensions of brand voice. 

The first live implementation was for a company called Brava Fabrics, where MS trained them on their product catalogue, brand guidelines, and tone of voice.

Overall — Marketing is at an inflection point, video is the headline act

The DM9 scandal is not just about bad judgment – it exposed the industry's fundamental unpreparedness. It asked questions about deepfake deception and where the line is drawn when video can conjure anything up with AI. 

Yes, video has been capable of lying for decades. But building a fake CNN news report with AI is far easier and more detached than filming it. 

Moreover, it highlights how companies might be shot down for ‘faking’ AI competence or missing the point. Authentic AI-enhanced marketing beats cheating.

THE STACK

Deep market research

AI deep research tools are seriously underrated for marketing and business use.

Market research, particularly, is time-consuming and can cost you tens of thousands of dollars if you hire agencies to do it for you. 

I used Claude to research the bubble tea market, and the results were extremely impressive, with 707 sources combed through in just 8 minutes, providing me with a wealth of amazing data and insights.

I have no business interest in bubble tea, but I know it’s an exciting, fast-developing industry.

The brief was simple: “I want to research the bubble tea market. We need to gauge the main market players and their market share and evaluate core players versus niche players. We will then identify gaps in the market based on our findings, taking into account both general market trends in general and customer needs. We will focus on the US.”

Pretty standard market research request. The kind that usually takes agencies weeks and costs thousands for a PowerPoint deck with insights everyone else already has. 

What went down

The initial research assessed over 300 sources in a few minutes, uncovering market data, the main players, and other macro-level information. What’s interesting here is the synthesis. The AI is connecting dots across different domains.

The next research I conducted focused on marketing, examining over 700 sources.

Claude linked gaming collaborations to cultural frameworks. It connected micro-influencer economics to social media strategy. It found platform performance data that helped explain regional marketing differences. 

Some successful viral campaigns came from gaming collaborations. CoCo Fresh Tea's partnership with Diablo IV launched themed beverages like "Frozen Orb" and "Lilith's Advocate" across 12 countries, complete with exclusive merchandise and in-game rewards.

Kung Fu Tea did something similar with Nintendo, creating limited-time tropical punch flavors for Pikmin 4 along with themed cups and exclusive straw caps. 

A welcome warning for bubble tea businesses

Here’s an interesting tidbit Claude surfaced – the October 2024 Dragons' Den "Bobba" controversy.

A Quebec company described traditional bubble tea as something consumers are "never quite sure about" that they were "making better."

When actor Simu Liu called out the problematic nature of "taking something distinctly Asian and making it better," the backlash was immediate and devastating.

Within six days, the investor withdrew support, and the company issued a formal apology. 

Based on this, Claude recommended that bubble tea businesses be cautious with authenticity when marketing their products – not bad advice at all. 

On the marketing front, the research recommended ASMR videos, behind-the-scenes content, and hyperlocal marketing. It recommended untapped markets, with secondary cities (100K-500K population) lacking adequate bubble tea options.

Claude also produce a few interesting case studies, such as:

• Tiger Sugar's ROI: Secured 1,873% return on ad spend (15.7x better than social benchmarks)

• Kung Fu Tea's digital ecosystem: 74% higher average orders per location

• OMOMO's strategy: Became an "influencer tea brand" with just 3 California locations

• CIA Bootleg Manila: “Psychological” research into Serinitea, discovering three brand pillars they used to drive sales across 60 stores. 

And here’s a nice insight about using micro-influencers, accompanied by some data and platform synthesis, again showcasing this method’s unique ability to tie together seemingly disparate sources.

Another interesting insight that surfaced was the demand for healthier bubble tea choices, which use stevia and agave to lower the sugar content. 

You get the idea!

I learned more about this market in 30 minutes than I could have ever without AI, including some quite niche factoids I might have never come across otherwise.

If anyone would like to view the full research, I’ve published it here. Note that I am not affiliated with Anthropic. ChatGPT does the job, but Claude 4 Opus is ridiculously good at this. 

The caveats

While synthesising results from hundreds of sources in minutes is incredibly powerful, there are some caveats.

First, you’ll need to be judicious about checking individual pieces of data. 

Second, it can be a bit too deterministic. For example, the research states that because TikTok's behind-the-scenes content generates 75% of impressions from the "For You" page for smaller accounts, that strategy must work for bubble tea. It’s a fair inference, but ultimately unproven.

Another example is the cultural appropriation point – that was just one incident, so while demonstrative of a potential risk, you’ve got to take it with a pinch of salt. 

Third, deep research consumes a lot of your context window, so you can’t keep working in that chat for too long before you need to start a new one. However, you can import your reports into a new project (just copy and paste). However, that does mean you lose the links unless you copy them out as well.

Overall, with some structuring, you could build all sorts of research-driven strategies using deep research.

This example barely scratches the surface. A few other uses I’ve personally experimented with include:

• Locating content ideas based on news reports/trending items, and stories in a given topic or domain

• Competitor analysis; looking at competitor content, blogs, social media strategy (slightly limited there as it’ll only find third-party data, but still decent)

• Mapping YouTube channels, their follower count, potential earnings, content types, etc

• Locating product trends and gaps in the market. As shown by the bubble tea research, consumers have concerns about the sugar content of drinks – it’s insights like that that drive new product lines

The takeaway: For any marketers who crave data and information, and want to be maximally informed, get stuck in with some deep research.

AI MARKETING EVENTS

🗓️ Recommended AI marketing events

Whether you’re building with AI, leading a marketing team, or just trying to stay sharp in a fast-moving space, events are still one of the best ways to plug in.

Great events do more than just deliver content — they give you context. You get to see what other teams are testing, what’s actually working, and where the market is headed.

They’re also one of the few places where you can meet like-minded people, talk shop with folks solving the same problems, and build a network that isn’t purely algorithmic!

Some top events to check out:

•  AI Agents Summit (Virtual) — September 18–19, 2025: Laser-focused on AI agents, copilots, and autonomous systems. If you're experimenting with automating parts of your marketing workflow or interested in agent-based design, this one’s worth attending.

• AI for Marketers Summit (Virtual) — November 13–14, 2025: Created specifically for marketing professionals. From prompt engineering to campaign automation, it’s a practical look at how AI is being used right now in real teams.

• Data + AI Summit (San Francisco) — June 9–12, 2025: Hosted by Databricks, there’s plenty of marketing relevance here — especially around LLM workflows, AI tooling, and future-proofing your stack.

• Ai4 2025 (Las Vegas) — August 11–13, 2025: A big-picture AI conference covering finance, healthcare, retail, marketing, and more.

Hope you enjoyed this week’s issue. If you missed the last newsletter, you can read it here.

If you found it useful, please recommend it to a friend or colleague.

Until next time. Happy marketing.

—The AI Marketer