July 24th-July 30th // Estimated Reading Time: 9 minutes

The most popular AI bootcamp in the marketing industry is BACK! The AI Accelerator helps marketers level up on all things AI for marketing and advertising so they can make a meaningful impact on their business, fast.

We’re thrilled to announce Blis, the T-Mobile Advertising Solutions omnichannel DSP, as our exclusive sponsor for this run!

Enroll now using promo code EarlyBirdsLearnAI for 20% off!

Meta and Google sign EU's AI transparency pact, while their ad tools keep going rogue
Source: Meta, Google
July 28th, 2026

Summary: Meta confirmed this week it's signing the EU AI Act Code of Practice on Transparency of AI-Generated Content. Google signed the same code last week. The code pushes companies to label AI-generated content and build in provenance tools, so users can tell what's real vs. AI. Meta touted the labeling work it started in 2024 and a new tool that checks if an image was made with Meta AI. Google points to SynthID, its watermarking tech, and partnerships with labs like OpenAI and NVIDIA to spread this tech industry-wide. The news comes as marketers say Meta and Google's AI tools are prone to adjusting ads🔒 and campaigns on their own, without marketer input, creating headaches. For example, Meta auto-enrolled REI into a feature that generated a bike ad with two handlebars, and Google's tool mistakenly used a pet vaccinations link for a livestock vaccine ad.

Opinion: The AI failed, it’s your fault, but there’s nothing you could have done about it? 🤔

Meta and Google aren’t incentivized to give marketers the control they need to prevent AI from messing up their campaigns. They are incentivized to automate and optimize as much as possible to generate more ad revenue. 

Marketers, don't wait around for Meta and Google to give you control. Do this instead:

Or just stop running ads through Performance Max or Advantage+!

AI slop is fooling the ad verification tools marketers pay for🔒
Source: Adweek
July 28th, 2026

Summary: Low-quality AI-generated content, appearing on sites created purely to capture ad dollars, is slipping past the ad verification tools marketers use to avoid junk ad placements, according to a new report from TAG, the ANA, and Fiducia. This kind of content now accounts for 1.3% to 2.4% of open web programmatic spending, and it's often outperforming quality publishers on metrics that marketers are looking at, including viewability and premium-quality scores, more than 70% of the time. CPMs on these sites topped out at $7.08, compared to the $6.15 average for legitimate content, meaning some brands are paying a premium to appear next to AI slop. Most AI slop inventory should fall under made-for-advertising (MFA) inventory category.

Opinion: Low invalid-traffic rates and strong viewability are supposed to be signals of quality, and in this case they're pointing brands toward junk instead. Oy vey. 

Verification tools need to AI-proof their technology. And marketers need to go back to basics: inclusion lists, direct-to-publisher deals, private marketplace buying, and better measurement.

Check out this free course we built in partnership with the ANA on programmatic supply quality and what marketers can do to optimize their ad supply.

StackAdapt's new AI hub moves marketers from goals to action 

What It Does: DSP StackAdapt's Ivy Studio is an AI hub built into its platform that lets marketers describe a goal in natural language and have AI agents handle the campaign forecasting and execution to achieve it. The company frames it as a shift away from dashboards, arguing that most ad platforms still force marketers to dig through menus and reports before taking action, while Ivy Studio moves straight from stated objective to action. Ivy Studio runs on Ivy, StackAdapt's AI engine, and the company says its own employees already use it for more than 15,000 AI-assisted workflows a week, with nearly 70% of that usage concentrated in its top 20 workflows.

Quick Take: Platforms will need to offer marketers both home-grown agentic solutions along with an MCP server marketers can plug their own agents into. StackAdapt has already done some of the latter, now it’s tackling the former with Ivy Studio.

Anthropic launches Claude Opus 5

What It Does: Anthropic is positioning Claude Opus 5 as nearly as capable as its top-tier model, Claude Fable 5, at half the price. Anthropic says Opus 5 does well on coding and knowledge-work benchmarks, but still trails behind Anthropic's most restricted model, Mythos 5, on cybersecurity tasks. It's now the default model for Claude Max subscribers and the strongest option available on Claude Pro.

Quick Take: This fits the broader industry shift toward competing on cost, especially as cheaper Chinese models put pressure on pricing from below. Marketers increasingly need to manage costs by optimizing which models they are using for their various AI workflows and tools. 

Channel Factory introduces AI Slop Detection

What It Does: Channel Factory's new tool sorts content into five tiers, ranging from confirmed AI slop on one end to fully human-made content on the other end, instead of the simple yes/no AI flag many detection tools use. It runs pre-bid at the channel, video, and creative level, across more than 50 languages, using image, audio, and keyword analysis. It covers YouTube, Meta, TikTok, Reddit, Snapchat, and CTV.

Quick Take: Giving brands a spectrum instead of a binary flag could be a more useful design for marketers, as AI content production becomes more nuanced and increasingly accepted by users.

Nielsen's ad data powers its new AI assistant

What It Does: Nielsen upgraded Ad Intel, its competitive intelligence product, with a conversational AI layer. Ad Intel already monitored 5.5M brands and 4.6M advertisers across 23 media types in more than 90 countries, covering everything from CTV to retail media to social. Now users can query that data directly and get real-time answers on competitor spend shifts, creative strategies, and emerging trends.

Quick Take: AI on its own = 🤷. AI + proprietary data = 🔥

The Setup: Michaels wanted to move shopping beyond traditional keyword search and give customers a way to describe their creative visions in their own words and get personalized guidance back.

The Solution: Michaels built "Ask Mike," a shopping assistant powered by Google Cloud's Gemini Enterprise for Customer Experience. Previously, when planning a kid's birthday party, a customer might search "balloons," then separately search for "party favors" and "cake toppers," piecing the items together themselves across several searches. With Ask Mike, a customer can just describe their project or event, the assistant asks a few follow-up questions, like theme, budget, age range, and returns a matched set of products and ideas covering the whole thing in one pass. Ask Mike launched in May. 

The Results:

Why It Matters: This isn't just a simple "AI beats search" thing. An AI assistant can generate tremendous value when customers face a complex buying decision instead of a single, obvious choice.

Your Action: Pull search logs for multi-query sessions, or cases where customers run several related searches in one visit. Then examine returns, exchanges, and support tickets for phrases like "didn't match," "wanted something different," or "problems with set-up." A frequent pattern is a business case for an AI-powered assistant.

Google quietly gives ground on PMax controls🔒 - Google is piloting a feature letting media buyers exclude what some consider to be junk inventory from Performance Max campaigns. The “black box” product is ever-so-slowly becoming more transparent.

Mark Zuckerberg says US should accelerate AI development, not restrict it🔒 - Zuck argued that optimism should be the default stance on AI, warning against restrictions and calling for any government model review to move fast.

Meta sinks 8%, Microsoft jumps 15%, as AI trade splits - Microsoft's stock had its best day since 2008 after Azure grew 43% and Copilot paid seats topped 30M, signs its AI spending is paying off. Meta shares fell nearly 8% after missing profit expectations and revealing plans to spend $130B-$145B on AI this year. Amazon (👍) and Google (👎) earnings were also heavily impacted by AI.

OpenAI's ChatGPT reaches the coupon stage of building an ad business🔒 - OpenAI is offering new advertisers promo credits, like $50 back after spending $50 within 14 days. The move comes as OpenAI builds out an ad sales org🔒 and ChatGPT nears 1B weekly users🔒 seven months past its target.

Time has started serving ads to AI agents🔒 - Time is working with startup Mobian to place sponsored brand FAQs into the stripped-down, text-only versions of its pages, which is what AI crawlers read. The pitch to brands: pay us to influence how you show up in AI responses.

Jeff Bezos wants to give Prime Video an AI makeover - Bezos reportedly wants to overhaul the Prime Video home screen, replacing its grid of tiles with a dynamic, AI-curated feed built around personalized recommendations rather than paid studio placements.

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!