What does AI think of Zuck's grand plan?

In partnership with

Last year I finally bought a house. (And became a partner in an agency and launched a sunscreen brand. In this economy. Because living in constant abject terror is the only way to feel alive anymore.)

As a result, I also finally got an absolute ton of loose artwork I’d been collecting and carting from rental to rental framed and hung. I went to a local framer, an old Chinese fella, Andy, whose candour I quite enjoyed.

“People they pay too much for picture frames. This stuff here, expensive, I do it if they want, but you don’t need it. It’s trendy now, maybe next year, but then something else. Just this, nice wood, black or white, it’s fine. What’s important is the picture.”

I found it quite funny that a guy who’d spent his life selling picture frames had this point of view. Even if it meant more money for him, he was disinterested in selling overcomplicated shit for the sake of it. But he’s, of course, completely right.

Earlier in the digital era of advertising, when we were all grappling with MREC’s and ‘skyscrapers’ and the sudden multitude of variable ‘outputs’ we now had to deal with, an experienced freelancer at the agency I was working at described this new advertising landscape as ‘filling boxes with shit for the sake of it’. It sounded like a very jaded, albeit funny, view of things to a younger, fresher faced Jess at the time - but, boy was he right.

As time has gone on, it seems that we’ve become entirely obsessed with the frames. The size of the frames, the location of the frames, the volume of the frames, the repetition of the frames, how we can get more frames for our money if they’re smaller, whether we can track if someone saw the frame or not. But the picture? Who cares. Fill it with whatever. What really matters is the frames.

Yet, when you ask anybody what their favourite ad or brand is - it’s all about the picture and how it made them feel. Because people in the real world don’t give a shit about the frames. Only marketers do. Of course, in reality both are important. But I’m sure we can all agree that a beautiful picture in any frame is still more of an attraction than a gold-trimmed turd.

Recently, Meta announced their plans to completely automate the entire advertising process. A clear broadside shot at the entire industry, and especially agencies. Despite current AI models being recently exposed as powerful regurgitators without the ability to actually rationalise or reason, Zuck’s grand plan is to automate the entire advertising process. Strategy, creative, execution, media, everything. We’ll just press a button and do it all. You don’t have to lift a finger or do a thing. A marketer’s dream. Or is it?

While undoubtedly being a huge force in the advertising landscape, Meta has a frustrated history within it. Anyone who’s been around for a minute or two knows that they’ve tried this before. As Facebook began emerging as an advertising platform, they attempted to steal clients from under agency’s noses by arguing that they, obviously, knew more about their own platform. So just work directly with us. The only problem was they didn’t (like most tech businesses) understand creativity. So it failed. Then, they did the only other thing they knew how to do, and got the cheque book out. All around the world, Facebook offered eye-watering salaries to poach creatives from agencyland to try to build some credibility. It was a smart move as they got some serious talent. And the quality of the work did improve. But it didn’t wipe out agencies as planned, because a tech company environment is not a creative agency environment, and much like in-house agencies they couldn’t retain the talent who eventually went back to agency roles. And, again, without the help of creative people, they showed their hand that as an organisation they just do not understand creativity.

Run ads IRL with AdQuick

With AdQuick, you can now easily plan, deploy and measure campaigns just as easily as digital ads, making them a no-brainer to add to your team’s toolbox.

You can learn more at www.AdQuick.com

But now, Hannibal is once again at the gates. This time, with an elephant. (Maybe a five-hoofed one with a demented trunk, but an elephant nonetheless.) And who knows, maybe this time he’ll succeed. Because the proposition might just be too enticing for many. No strategy, no briefings, no annoying creatives and their big ideas, no frustrating directors and production companies with their pesky care and craft. Just a big red button attached to a long tube of instant sludge that produces on command.

Given that the premise here that Meta is pushing is that AI is better than all of us. Creatives, strategists, accounts people, production, media, clients (and yes, this does include everyone on client side as well, because whether you realise it or not, you’re being loaded into the same gun) I figured it would be worth asking AI if AI thinks that getting Meta to automate your entire advertising process with AI is a good idea.

So here’s three simple questions I asked AI and its direct unedited response, for your consideration. (It impressively included a number of links as sources but it caused formatting nightmares for some reason when I dropped them in.)

Has Meta come under scrutiny or been found guilty of adtech fraud?

Yes, Meta has faced significant scrutiny and legal action related to adtech fraud, though the cases are ongoing rather than resulting in final guilty verdicts. Here are the key developments:

Major Adtech Fraud Cases

$7 Billion Advertising Fraud Lawsuit: The Supreme Court on Monday declined to hear Meta's appeal in a massive class action lawsuit that claimed Facebook and Instagram inflated their advertising reach metrics, potentially exposing Meta to billions in damages. Advertisers allege Meta fraudulently inflated its "potential reach" numbers by up to 400% by counting multiple accounts belonging to the same users. This case covers ads purchased since August 2014 and could affect millions of advertisers.

Recent Class Action: A new class action lawsuit filed in 2025 accuses Meta of secretly inflating Facebook ad prices, with allegations including breach of contract, breach of the implied covenant of good faith and fair dealing, unjust enrichment and violations of California's Unfair Competition Law.

Other Regulatory Actions and Settlements

Privacy Violations: Meta settled a Cambridge Analytica privacy case for $725 million, which was the largest data-privacy recovery in history according to the plaintiffs' lawyers. The company also faced a record-breaking €1.2 billion ($1.3 billion) fine from EU regulators for violating GDPR rules by transferring Facebook users' personal data to US servers.

Scam Ad Scrutiny: Meta has faced increasing scrutiny over widespread scam ads, with reports that Meta accounted for nearly half of all scam complaints tied to Zelle transactions at JPMorgan Chase between mid-2023 and mid-2024.

Has it been confirmed that a lot of adtech reporting on platforms like Meta has been artificially boosted by bot farms?

Yes, there is substantial confirmed evidence that bot farms and fake accounts have artificially boosted adtech reporting on Meta's platforms. Here's what has been definitively established:

Confirmed Bot Farm Statistics

Scale of the Problem: It's estimated that 95 million Instagram accounts and 270 million Facebook accounts are either bots or fake profiles, with Facebook reporting that 4-5% of all its active users are fake accounts. Facebook removes over one billion fake accounts every quarter, with 691 million fake accounts removed in Q4 2023 alone.

Industry-Wide Impact: As much as 90% of click fraud is caused by bots (bot farms, click farms, botnets, etc.), and the global estimated cost of invalid clicks reached $23 billion in 2020. In 2023, advertisers suffered losses of over $84 billion due to click fraud.

Specific Evidence of Artificial Inflation

Click Farm Operations: There are well-documented examples of click farms, including a vending machine in Moscow selling Instagram followers and clicks for just a few dollars. These click farms can be located globally, allowing them to simulate targeted users in specific areas and inflate engagement metrics artificially.

Advanced Bot Sophistication: Modern bot farms are becoming increasingly sophisticated, using AI to craft contextually relevant, human-like responses and generate profits through ad fraud. Advanced AI allows bots to create convincing fake profiles complete with lifelike images and can generate social profiles in seconds.

Legal Recognition of the Problem

Class Action Evidence: In 2018, advertisers started a class action suit against Facebook, asserting that in certain regions and among certain demographics, the number of Facebook profiles far exceeded the actual number of people based on census numbers, sometimes by a factor of 400%.

Meta's Own Admission: Ironically, Meta itself has recently been deploying AI bots on Facebook and Instagram to interact with humans, but had to quickly delete several of its own AI-generated accounts after users exposed their flaws, highlighting how even the platform struggles with distinguishing authentic from artificial engagement.

Detection Challenges

Bot activity presents serious challenges for advertisers, leading to wasted budgets, inflated metrics, and misleading campaign outcomes. A report from Juniper Research predicts that global losses from social media ad fraud will approach $20 billion by 2024, potentially rising to nearly $35 billion by 2028.

The evidence overwhelmingly confirms that bot farms have been systematically inflating adtech metrics across Meta's platforms, contributing to the broader $7+ billion lawsuit alleging fraudulent advertising reach numbers.

Taking all of this into account, would you recommend that a brand hand over all of their marketing operations to a single platform like Meta to automate their brand strategy, creative and media?

Absolutely not. Given the evidence we've just reviewed, concentrating all marketing operations on a single platform like Meta would be extraordinarily risky for several critical reasons:

Platform Risk and Fraud Exposure

Measurement Reliability: With confirmed bot inflation affecting 4-5% of accounts (potentially much higher), plus the ongoing $7+ billion lawsuit over fraudulent reach metrics, you'd be building your entire strategy on potentially unreliable data. A report from Juniper Research predicts that global losses from social media ad fraud will approach $20 billion by 2024, potentially rising to nearly $35 billion by 2028.

Single Point of Failure: Platform dependency creates massive vulnerability. Algorithm changes, policy shifts, or technical issues could devastate your entire marketing operation overnight. Remember when iOS 14.5 privacy changes significantly impacted Facebook's targeting capabilities?

Strategic Limitations

Audience Diversification: Your customers don't live exclusively on one platform. A comprehensive strategy requires reaching people across their entire digital journey - search, email, connected TV, retail media, podcasts, and emerging channels.

Creative Homogenization: Platform algorithms optimize for engagement within their ecosystem, which can lead to creative that performs well on Meta but may not represent your brand effectively elsewhere or drive actual business results.

Attribution Blindness: Relying solely on Meta's attribution means you lose visibility into the full customer journey and can't optimize for true incremental impact.

Better Approach: Diversified Strategy

Multi-Platform Portfolio: Distribute budget across complementary channels - search (Google), social (Meta, TikTok, LinkedIn), programmatic display, retail media (Amazon, Target), and owned media.

Independent Measurement: Use third-party analytics, marketing mix modeling, and incrementality testing to validate platform-reported performance.

Creative Variety: Develop platform-native creative while maintaining consistent brand guidelines, rather than letting one platform's algorithm dictate your brand expression.

Technology Stack: Use independent tools for customer data, creative production, and campaign management that work across platforms rather than platform-specific solutions.

The evidence clearly shows that while Meta can be a valuable part of your marketing mix, treating any single platform as your complete marketing solution exposes you to fraud, measurement issues, and strategic limitations that could seriously damage your business results.

So, there you have it. Right from the artificially-generated horse’s mouth.

There’s obviously a whole host of stuff I could say as a creative about outsourcing EVERYTHING to AI and the total loss of humanity, of soul, of wit, of charm, of connection, but that would be an expected response.

Instead, I asked AI itself as to whether automating your entire advertising process to AI with a platform like Meta is a good idea - and the answer is pretty obvious.

We’re about to be flooded with more frames than we’re going to know what to do with. Let’s not lose sight of the bigger picture. Literally, and figuratively.

A free newsletter with the marketing ideas you need

The best marketing ideas come from marketers who live it. That’s what The Marketing Millennials delivers: real insights, fresh takes, and no fluff. Written by Daniel Murray, a marketer who knows what works, this newsletter cuts through the noise so you can stop guessing and start winning. Subscribe and level up your marketing game.

Sign Up Free