July 3rd-July 9th // Estimated Reading Time: 7 minutes

Reddit is cracking down on AI marketing slop with its own AI
Source: Bloomberg
July 6th, 2026

Summary: Reddit's AI detection systems flagged 25K spammy posts and comments per day in the first quarter, cutting user exposure to spam by 20% compared to a year earlier. A chunk of that comes from brands seeding posts meant to look like genuine opinion, in hopes that AI chatbots like ChatGPT or Gemini will pick it up and spit it back out to users as recommendations, since Reddit ranks among the most-cited sources in AI responses. Reddit isn't alone in going after AI slop. Google reportedly took down 50K AI slop networks publishing low-quality, AI-generated video in coordinated, suspicious patterns. Meanwhile, the ANA's latest programmatic benchmark report found that ad spend flowing to made-for-advertising (MFA)🔒 sites rose for the first time since 2023, a reversal the report links partly to the growth of AI-generated content.

Opinion: There are two simple takeaways here for marketers.

Takeaway #1: Don’t create shitty AI marketing content (aka “AI slop”). Sure, it may seem like it’s filling gaps at first and helping you scale, but eventually it’ll start making your brand look bad, it’ll get flagged, it’ll get taken down, and it'll end up being a huge waste of time and resources. The Reddit case is a warning for any brand or agency scaling content production with AI without human editorial judgment behind it.

Takeaway #2: Don’t let your ads run on shitty content. The Google story and the uptick in MFA are warnings about where your ad dollars could end up in the era of cheap AI content production. It’s bad for your brand and it doesn’t perform. (Check out this free U of Digital course we built in partnership with the ANA on how to optimize ad supply quality.)

Meta unveils Muse Image and Muse Video

What It Does: Meta released its first image generation model, created by its Meta Superintelligence Labs division. The tool can make precise edits based on natural language instructions and blend several source photos into a single output. It can run a web search to pull facts or images for reference and write code to generate precise outputs like QR codes. It's rolling out now inside the Meta AI app and site, Instagram Stories in the US, and WhatsApp for select regions, with Facebook access on the way. Advertisers get access too, through Meta's Advantage Plus, where the model can remix existing ad assets into new, on-brand versions. Meta also previewed Muse Video, a video AI tool with synced audio, coming later to creators and Meta AI.

Meta also launched Muse Spark 1.1, its agentic and coding AI model that it is charging users for. With its launch, Meta has officially joined the frontier model race and is making up ground. Mark Zuckerberg is on a quest to bring down the cost of AI.

Quick Take: Meta used to rent image generation and video AI tech from Midjourney and Black Forest Labs. Owning the model gives Meta more control over the entire ad pipeline, from targeting data to the creative. And like Mark said, it lowers the cost. That's a potential structural advantage, though Meta still has a long way to go in order to prove it can compete with Google, OpenAI, and the other frontier AI model companies.

OpenAI launches GPT-Live, a new voice model for ChatGPT

What It Does: GPT-Live changes how ChatGPT handles spoken conversations. Earlier versions needed you to stop talking before they'd respond, which made exchanges feel choppy and unnatural. This version can listen and talk at the same time, so it can chime in the moment or hold back if you're still working through a thought, more like how humans actually converse. It’ll even say “mmhmm” while you’re talking to keep the conversation flowing.

Quick Take: ChatGPT’s voice functionality has been strong, but you can still tell you’re talking to AI because of the weird delay or awkward pause before it answers. This should take it to the next level. High-quality AI voice recognition and chat is truly a game changer and still feels like AI’s best-kept secret. Marketers should be all over this! (e.g. conversational audio ads, anyone?!)

The Setup: Conair's e-commerce team, which oversees subsidiary Cuisinart, had a backlog of ad videos stuck in production for three to six months at a time. The company wanted to keep up with growing video engagement on Amazon without waiting months per asset.

The AI Solution: After seeing Amazon's Creative Agent demoed at an Amazon Ads conference last November, Conair ran a 30-day A/B test with agency Global Overview to produce a 15-second Cuisinart food processor video. The agency fed Conair's creative briefs into the AI tool, which generated roughly 80% of the finished video. Human editors handled the rest, fixing logos, colors, and other brand details the AI got wrong.

The Results:

• The AI-produced video took about four weeks to complete, compared to three to six months for a traditionally produced video.

• The video drove 18% higher detailed page views versus a traditionally produced video.

• Cost per detail page view dropped 14%.

Why This Matters: The video wasn't fully automated, but the 80/20 split between AI output and human polish was enough to collapse a months-long production cycle into weeks, all while outperforming the traditional version on engagement.

Your Action: Test the efficacy of AI video on something with zero video coverage. Build a detailed brief for the AI tool the way you would for an agency, and keep a human review step for brand-specific details like logos and colors. Launch it and monitor performance. If you can generate on-brand video, and there’s uplift in performance, ramp up production everywhere you don’t have video coverage immediately. This is low-hanging marketing fruit!

OpenAI exec Fidji Simo says she’s stepping down due to chronic illness - The former Facebook and Instacart exec went on medical leave three months ago and says her recovery will be "much longer and more complex" than anticipated. She will transition to a part-time advisor role. ​​

ChatGPT is making room for ads from regulated verticals—but marketers aren’t ready to jump in at scale - OpenAI loosened its ad policy in April to allow case-by-case approval for financial services, healthcare, and legal advertisers, but agencies say finance clients remain cautious. Non-pharma healthcare brands are pushing to get in. 

These creators are hacking ChatGPT to rake in brand deals🔒 - Creators surfacing in AI chatbot answers are seeing real payoff. One fashion creator's brand inquiries jumped 50% when ChatGPT and Claude began recommending him.

Boostr and Vox Media execute agentic media buy - Boostr and Vox Media completed a media buy with zero manual intervention. AI agents negotiated budget, audience, and delivery terms while humans only verified and launched the campaign.

Meta adds updated disclosure tags for AI-generated ads - Facebook and Instagram will now flag when ads are AI-generated, whether the content came from Meta's own generation tools or an outside program like Photoshop or Dall-E.

New York Times and other news outlets accuse OpenAI of lying in discovery, demand legal sanctions - Publishers want OpenAI sanctioned for allegedly telling the court it couldn't search its own training data and logs for two years, then deleting logs once a deposition proved it could.

Dollar Shave Club's bet: AI makes agencies optional, not obsolete🔒 - The brand handles 90% of its ad production in-house already, and its brand chief wants AI to increase that percentage. Hopefully

The U of Digital AI Edge Newsletter is intended for AI Edge Pro subscribers, but occasional forwarding is okay!

To subscribe visit Uof.Digital/Newsletters/ai-edge or AI Edge Pro, or contact us directly for group subscriptions.

And remember, U of Digital helps teams drive better outcomes through structured education on critical topics in marketing / advertising like AI, mar tech, programmatic, CTV, commerce media, and more. Interested in learning more about how we can supercharge your team? Let’s talk!