July 10th-July 16th // Estimated Reading Time: 10 minutes

OpenAI’s first hardware device is reportedly a screenless speaker that can move
Source: TechCrunch
July 14th, 2026

Summary: Bloomberg reports that OpenAI's first hardware product will be a “screen-less smart speaker that can move” with a camera and sensors built-in. Internally, people describe it as a companion-like device meant to get to know its owner and draw on personal data such as email. Apparently, the device also involves “mechanical elements that can move on their own.”

Former Apple engineers who helped build the iPhone and Mac worked on it. As OpenAI focuses on hardware, it is shutting down its Atlas browser, folding its AI browsing features into a Chrome extension and the ChatGPT desktop app instead. 

Opinion: News of an OpenAI device built by ex-Apple engineers leaking days after Apple sued OpenAI for stealing its trade secrets is … not a good look.

Lawsuits aside, we believe one of the major step changes in how people use technology will revolve around using voice instead of writing / typing. While Alexa and Google Home devices have pushed us in this direction over the years, their voice technology has been too clunky to meaningfully change how consumers behave. All we can trust Alexa to do is set a timer or tell us the weather. OpenAI’s voice technology is far superior.

Today, OpenAI’s voice technology lives inside the ChatGPT app on your phone or in a browser window. Those environments aren’t conducive to behavioral change; people are used to staring and typing, so using voice doesn’t feel like the natural, first option. Imagine a device that is always around you, that can listen and speak to you, can be your eyes and ears, is connected to everything in your life (your data, your devices, your home, etc.), and is constantly working on your behalf. No screen, no typing, only voice. A true “AI companion” of sorts. If there are killer use cases of such a device (there are) that the masses will gravitate to (they may), that could serve as the catalyst for this step change. And if that step change happens, it’ll introduce a whole new world for marketers.

Check out this this clip of Sam Bloom, Head of Partnerships at PMG, talking about how the search box will soon become the “voice box” on this week’s AI Edge Podcast.

Google Images gets a redesign and AI Overviews can now generate images

What It Does: To mark Google Images' 25th anniversary, Google is rolling out a new browseable home for Images with a real-time, personalized gallery you can save to collections. Google has also added AI image generation directly to AI Overviews using its Nano Banana model, which can turn a text prompt into a custom visual. The new Google Images home starts rolling out over the coming weeks on desktop in the US.

Quick Take: Google is turning Images from a search results page into an interactive, AI-driven discovery feed, and letting people generate the image they can't find instead of settling for one that already exists. Could marketers take a page from this playbook by letting their customers ‘conceive’ new product designs using AI if they don’t find something they like in their catalog?

ChatGPT Ads adds custom audience targeting

What It Does: OpenAI's ChatGPT Ads now lets advertisers upload their own customer or prospect lists, using emails or phone numbers, to target campaigns. Advertisers can include or exclude specific audiences from a campaign, or set bid multipliers to spend more aggressively on high-value segments. Each audience needs at least 25K matched users before it can be used, and OpenAI recommends 100K or more.

Quick Take: This puts ChatGPT Ads’ 1st party data targeting capabilities on par with Meta and Google. That said, “retargeting existing customers on AI” could be a huge unlock. Let’s say you sell shoes online, a user visits your site and puts shoes in their cart, but never converts. Now imagine that user, weeks later, goes to ChatGPT and has a random conversation about packing for a trip. ChatGPT could theoretically still connect the dots back to that cart and show the user an ad for that product in the perfect moment that gets them to buy.

Warner Bros. Discovery's new ad platform uses AI agents to unify linear and digital buying

What It Does: WBD rebuilt its ad platform on AWS to let advertisers plan and buy across its linear and digital inventory from one place instead of separate systems. AI agents can recommend which slots to allocate based on campaign goals, then adjust those recommendations during the campaign. WBD says the platform will also make it easier to run newer formats like interactive ads. Unified media planning arrives in Q3, with tools for managing orders, pricing, and campaign delivery following in Q4.

Quick Take: Add WBD to the growing list of ad platforms handing buying decisions to AI agents. Marketers should be wary of pre-built agents that live in ad platforms: you can’t be sure if they’re optimizing for your interests or the platform’s.

The Setup: Bayer's media team wanted to relieve their pain of losing customers to competitors and wasting ad dollars chasing the wrong prospects. Standard programmatic targeting treats audiences as broad cohorts, which can make it difficult for brands to tell which individual customers are worth pursuing.

The AI Solution: Bayer links its sales data with audience data in real time so it can flag customers likely to churn and reach them with retention messaging before they do. On the growth side, the team is using AI to find existing and would-be category buyers, drawing on shifts like the rise of GLP-1 drugs to spot where new demand is forming. Bayer is using AI to set a different bid for each individual shopper based on how likely that person is to buy a product, as opposed to bidding the same amount for an entire audience. Bayer is also piloting contextual AI in connected TV to figure out which specific shows tend to drive the best results.

The Results:

• 4x better targeting than standard programmatic targeting in early tests

• Gains in both ROI and broader business outcomes

• The predictive targeting approach is now being rolled out to more brands across Bayer's portfolio

Why This Matters: Marketers have always wanted to know which customers are about to leave, but that intel usually arrives too late to act on before they churn. Bayer's AI model can get there in advance, turning a problem marketers have lived with for years into something they can finally get ahead of.

Your Action: Pull 12 months of churn data and feed it into a predictive AI tool to identify which signals most strongly predict when a customer is about to leave, whether that's a support complaint, a missed reorder window, or a drop in app activity. Use the tool to score customers on churn risk in real time. Start small: Apply the scoring to your highest-value customer tier, build a retention campaign that triggers off high-risk scores, and run it against a holdout group. Compare churn rates between the two groups after 60 days. If the experiment works, expand it to your next customer segment.

Meta's AI ad tools keep botching brands' creative🔒 - Advertisers say Meta's AI features are randomly turning on and mangling ads, such as REI's viral two-handlebar bicycle. Meta says advertisers are responsible for reviewing AI outputs before they run.

Meta expands its ads MCP server with open access and new guardrails - On Thursday, Meta opened its ads MCP server to any developer with a Meta app, and added rules letting business admins control what AI agents can do on their ad accounts, from managing budgets to catalog updates.

AI videos are taking over TikTok Shop affiliate videos🔒 - Creators are using AI-generated twins to demo products. SharkNinja has banned the practice, pulling commissions from affiliates who use AI content. TikTok allows it platform-wide with disclosure.

Meta AI can now use your Instagram photos without consent - Meta's Muse Image model lets anyone generate AI images using any public Instagram account's photos, just by tagging that account in a prompt. Users can opt-out of their photos being included. An actors union and talent agency are pushing back on the default opt-in.

OpenAI's ad forecast may be off by 90%, analyst says🔒 - OpenAI projects $100B in ad revenue by 2030, but EMARKETER puts the entire US chatbot ad market's ceiling at just $5.41B by then. A separate analyst model is more optimistic, forecasting a bull case of $110B and a bear case of $24B. In any case, OpenAI’s forecast is inflated.

Most Americans won't pay for AI tools, survey finds🔒 – 96% of Americans have tried AI and 86% use it routinely, but only 26% currently pay for a platform. More than half said they'd drop a tool if it required payment. 

Reddit questions if AI data deals could hurt its ad business🔒 - A Reddit ad exec acknowledged an internal debate over whether licensing its community content to OpenAI and Google for training chips away at the data advantage that differentiates Reddit's own ad business. This is now the existential question for any publisher, with seemingly no right answer.

AI search cut publisher ad supply by up to 40% in Q2🔒 - As AI search sends less traffic to news sites, publisher ad requests dropped 32%-37% in the US and 39%-41% in the UK, per Ozone data. Rising ad prices are cushioning the blow for now, with eCPMs up 30% in the UK and 7% in the US.

Why AI visibility now depends on earned media - A new Mediassociates guide finds more than a third of consumers start their searches with AI tools, and 60% of searches end without a click. AI engines favor third-party coverage over brand-owned content by a wide margin.

Anthropic's newest ad is creeping people out - The spot is titled "There's hope in hard questions." Our hard question for you: Do you like this ad?

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!