Top Retail News You Need to Know This Week

This week’s retail news shows a market adjusting to three pressure points: AI monetization and trade policy uncertainty.

• OpenAI tests ads in ChatGPT with major retailers: OpenAI is running its first ChatGPT ad pilots with top retailers, using shopping as a test for AI monetization. By driving purchase decisions, it aims to prove AI’s real-world value and push brands to track AI-driven discovery and conversions across the full funnel.
• Tariff relief uncertain as price pressures build: The US Supreme Court limited the administration’s emergency tariff authority, potentially allowing refund claims though recovery will depend on legal deadlines and documentation, not on whether costs were previously absorbed or passed on.. Meanwhile, Amazon CEO Andy Jassy says the tariff “buffer period” is ending, meaning sellers who have been absorbing higher costs may begin raising more prices noticeably in 2026.

These developments point to a more disciplined approach in 2026, and we unpack what that means in more detail below.

Conversational Commerce Begins: Retail Tests AI Ads

The next phase of retail media isn’t happening on a marketplace homepage or a Connected TV screen, it’s unfolding inside a chatbot.

OpenAI has started testing ads in ChatGPT for logged-in adult users on its Free and Go tiers in the US, while Plus, Pro, Business, Enterprise, and Education plans remain ad-free. The rollout is limited for now, giving the company time to gather feedback and refine the experience before expanding.

But this isn’t just a small feature update, it marks a fundamental change in how people may discover products.

How ChatGPT Ads Work, And Why It Matters

OpenAI is positioning ads as supportive, not intrusive. Several guardrails define the pilot:

• Ad placement and visibility: Ads appear below responses, clearly labeled and separated from answers. They won’t show in Temporary Chats, when logged out, after image generation, or in the ChatGPT Atlas browser.
• Answer independence: Ads do not influence ChatGPT’s responses and are clearly labeled and visually separated.
• Privacy protections: Advertisers do not see chat logs or personal data. They receive only aggregate performance metrics (e.g., impressions, clicks).
• Sensitive-topic exclusions: Ads won’t appear next to content involving health, mental health, politics, or for users under 18.
• Cost-per-view model: Advertisers are charged on impressions.
• User control: Free-tier users can opt out of ads in exchange for fewer daily messages and can manage personalization settings.

The strategy is pretty straightforward. Ads are meant to help fund broader access to AI infrastructure, while keeping the paid tiers completely ad-free.

For businesses, OpenAI has hinted that more ad formats and buying options will roll out over time. This isn’t just a one-off experiment, it’s the early stages of what could become an entirely new media channel.

Retailers Step In Early

The first wave of participants includes:

• Target
• Adobe
• Williams-Sonoma
• Albertsons

These retail giants aren’t experimenting casually. They’re positioning themselves at the point of intent. For example:

• Albertsons tested Valentine’s Day ads that appeared alongside queries like “best flowers for Valentine’s Day.”
• Target is leveraging its Roundel retail media network to meet customers during conversational shopping moments.
• Williams-Sonoma is exploring contextual ad placement while consumers compare product options.

Consumer attention is clearly changing. If consumers start using ChatGPT to decide what to buy, ad budgets will follow that behavior. It’s likely that other major AI platforms like Gemini will introduce their own ad product soon, turning conversational AI from a simple tool into a new retail surface where discovery and spending happen.

Why Retail Is the Real Test for AI

This isn’t just about ad inventory. It’s about validation.

Retail sits at the center of a $30 trillion global industry. As Microsoft CEO Satya Nadella has warned, AI valuations depend on benefits extending beyond tech companies. If non-tech industries, especially retail, don’t see measurable upside, bubble concerns grow louder.

OpenAI CFO Sarah Friar has made “practical adoption” a 2026 priority, specifically helping consumers decide what to buy.

And that’s the key distinction.

AI has largely been deployed behind the scenes in retail, e.g., optimizing supply chains, inventory forecasting, and employee workflows. But scaling AI in customer-facing experiences is proving much harder.

So, retailers are uniquely positioned to prove whether AI works in high-volume, consumer-facing environments where trust, accuracy, and brand control matter.

Amperity’s 2025 survey shows that while 45% of retailers say they’re using AI, only 11% feel ready to scale it. Fewer than half are deploying AI in customer-facing applications like personalization or chatbots, and just 23% are focused on preparing and connecting customer data for marketing purposes, the foundation AI depends on. More than half of respondents cited fragmented data (or data scattered across different systems) as their biggest barrier to scaling AI. Others pointed to high costs and limited technical expertise as additional challenges.

Retail may be enthusiastic about AI, but structural readiness is lagging. And because retail operates at a massive scale and directly touches consumers, the margin for error is slim. In high-volume environments, even small data gaps or flawed recommendations can quickly erode trust and damage brand credibility.

Imagine a shopper asks a chatbot, “What’s the best hypoallergenic skincare for sensitive skin?”

If the AI recommends a product that actually contains common irritants, and the customer has a bad reaction, that’s not just a software error. The shopper may blame the brand, the retailer, or the platform, and lose trust in all of them.

The Monetization Dilemma

There’s also a business model pivot underway.

OpenAI currently generates roughly 75% of revenue from subscriptions. Analyst Nikki Baird predicts a stronger shift toward B2B monetization, advertising in particular. But monetizing both consumers and advertisers simultaneously is delicate.

If ads feel intrusive or compromise trust, usage declines. If ads feel invisible, revenue lags.

Competitors like Anthropic have leaned toward consumer positioning, even critiquing ad-based approaches. Meanwhile, Microsoft Copilot and Perplexity have experimented with formats, but success metrics remain unclear.

The industry is still in early innings.

Retail Media’s Measurement Problem Just Got Bigger

Here’s where the story becomes more complex.

Retail media spending keeps rising, with in-store media (digital advertising that appears within a retailer’s physical location–think ads on smart carts) alone projected to surpass $1 billion by 2029. Yet measurement hasn’t kept pace.

According to AdExchanger, most retail media networks still rely heavily on:

Last-click attribution, short-term ROAS metrics, and most critically disconnected onsite  versus in-store reporting, “onsite” referring to a shopper’s digital activity on a retailer’s website — for example, clicking a sponsored product ad versus “in-store” where the actual transaction might happen at the physical register.

The problem: these two data streams often don’t talk to each other.

If a shopper clicks a sponsored ad on Target.com and later purchases that product inside a Target store, the digital team may never see the sale attributed to the ad — and the store team may never see the digital influence behind the purchase. The result is underreported ad impact, misaligned incentives, and optimization decisions based on incomplete revenue visibility.

Now add conversational AI into the mix.

If a shopper asks ChatGPT to compare meal kits, clicks an Albertsons ad, researches further, then buys in-store three days later, how is that attributed?

AdExchanger’s call for full-funnel measurement is where the path forward begins. 

In a stronger setup, the shopper clicks the ChatGPT ad while logged in, tying the interaction to a known customer ID. If that same customer later purchases in-store using a loyalty account, the retailer can connect exposure to transaction through closed-loop attribution.

Conversely, based on how purchases are often attributed, the ChatGPT click is tracked and the in-store purchase is tracked separately, but there’s no reliable way to connect the two. As a result, the sale might be credited to organic traffic, in-store merchandising, or the last digital interaction before purchase (or not attributed to media at all). That’s where measurement gaps begin to surface.

To truly evaluate ChatGPT ads, for instance, retailers must connect:

• AI-driven discovery traffic
• Onsite engagement
• Loyalty program signals
• In-store transactions
• Repeat purchase behavior

That’s why longitudinal– tracking customer behavior across multiple interactions over time – privacy-safe first-party data becomes the backbone of credibility in retail media.

Consumers are still spending, but many brands struggle to show how their ads influence the full journey from first exposure to final purchase. When that connection isn’t visible, teams end up optimizing for easy-to-see metrics like clicks or impressions instead of focusing on what actually brings in new customers and real sales growth.

What This Means for Retailers

1. Discovery is moving upstream

AI assistants influence decisions earlier in the buying journey.

2. Intent data is becoming pricier

Conversational context provides high-value signals that retailers may not fully own. If a conversation happens inside ChatGPT, retailers may only see clicks or traffic later, but they may not fully own or access the original conversational context that revealed the shopper’s intent. So, retailers are left to use their own customer data to build profiles for targeting.

3. Retail media expands beyond retail sites

The boundary between “platform” and “publisher” is blurring. Traditionally, retail media meant ads shown on a retailer’s own website or app like a sponsored product on Walmart.com or a promoted listing on Amazon, where the retailer acted as both the shopping platform and the ad publisher. Now that’s changing. If a shopper asks ChatGPT for the best espresso machine and a retailer’s ad appears in that conversation, the ad is running inside an AI platform, not the retailer’s site. That means discovery can begin outside retail properties, and platforms like ChatGPT start acting as publishers in the shopping journey.

4. Trust becomes a competitive advantage

Brands participating early must balance visibility with consumer confidence. If the ad feels pushy or irrelevant, it can quickly damage confidence in the brand. In conversational AI, trust becomes part of the buying experience itself.

5. Measurement infrastructure is now strategic

Without unified, closed-loop data, conversational ads risk becoming another black box, where money is spent but real impact is unclear. If a shopper clicks a ChatGPT ad but later buys in-store without a loyalty link, the sale and the ad may never be connected.

Overall, ChatGPT’s ad pilot isn’t just about generating new revenue. It’s a real-world test of whether conversational AI can work as a significant retail channel. The brands that win won’t just be the first to participate, but the ones that can turn shopper intent into actual sales, maintain trust, and clearly measure impact from start to finish. Retail is where AI now has to prove it delivers real results, not just big ideas.

Trump Tariffs Ruled Unlawful, Retail Still Faces Rising Costs 

Retailers may have more clarity on tariff authority, but cost pressure isn’t going away. Refunds will take time, new tariffs are emerging, and Amazon warns the inventory buffer is ending. In 2026, retailers must manage margins in real time, not wait for relief.

US Supreme Court Tariff Ruling: Legal Clarity, Operational Pressure

On February 20, 2026, the Supreme Court ruled 6–3 that the administration exceeded its authority under the International Emergency Economic Powers Act (IEEPA) when imposing broad global tariffs without congressional approval. The decision invalidates a key legal pathway used to apply sweeping tariffs.

For retailers and supply chain leaders, this is not simply a constitutional footnote. It creates immediate operational implications.

What changes now:

• Clearer limits on executive tariff authority
• Potential eligibility for duty refunds
• Greater scrutiny on compliance documentation

The Treasury has collected over $133 billion in import taxes under the emergency powers framework, with some estimates running far higher. Companies that paid duties under the now-invalidated authority may be able to pursue refunds through US Customs and Border Protection (CBP). More on that below.

For consumers, unfortunately, any potential refunds would primarily go to importers since tariffs are legally treated as payments made by businesses, even if those costs were later passed on through higher prices.

How To Get Tariff Refunds

How retailers pursue refunds will depend largely on whether their entries (shipments of goods imported into the US) are liquidated (CBP has finalized the duty assessment on the shipment) or still pending.

A significant share of IEEPA-related imports remain unliquidated. As of December 2025, roughly 19.2 million of 34 million affected entries had not yet been liquidated. For those entries, refunds could be issued automatically at liquidation if the tariffs are ultimately deemed unlawful. The Treasury has statutory authority to return excess duties when more was paid than required by law.

For recently liquidated entries, CBP may voluntarily reliquidate within 90 days to correct legal or classification errors. After that window closes, importers generally must file a formal protest within 180 days of liquidation to preserve refund rights. Timing isn’t just procedural, as it directly determines whether a refund is possible. That’s why some law firms advised companies to file protective protests while waiting for the Court’s decision in order to preserve their right to a refund in case the tariffs were struck down.

If protests are denied, or if protestability itself is challenged, companies may seek relief in the US Court of International Trade. The court has authority to review tariff disputes and, in some cases, order reliquidation and refunds.

The path exists. But it is time-sensitive and heavily documentation-driven, reinforcing that tariff recovery is an execution discipline, not an automatic outcome.

To qualify, companies must provide:

• Entry summary documentation
• Accurate Harmonized Tariff Schedule (HTS) classifications
• Proof of duty payment
• Transaction-level reconciliation
• Strict adherence to filing timelines

In addition, Justice Brett Kavanaugh noted the refund process may be complicated and take 12-18 months, and for companies with fragmented systems, it will be.

Tariffs don’t live in one system. They affect:

• Item setup and HTS codes
• Country-of-origin and Importer of Record status
• Purchase order cost assumptions
• Shipment documentation (e.g., Advanced Shipping Notice, routing, and freight)
• Invoice reconciliation
• Inventory valuation

If this data is scattered across Enterprise Resource Planning (ERP) systems, customs brokers, warehouse management systems, and accounts payable platforms, refund recovery becomes manual and error-prone.

Common failure points often include SKU-level misclassification, mismatched landed cost assumptions between partners, incomplete ASN documentation, and customs entries that don’t reconcile cleanly with purchase orders. The ruling highlights an uncomfortable reality that tariff exposure isn’t just a policy issue, it’s fundamentally a data integrity problem.

Tariff Relief Won’t Immediately Ease Margin Pressure

At first glance, invalidating tariffs should reduce costs. But several factors complicate that assumption: 

1. New Tariffs are Already in Motion

Within hours of the ruling, the administration announced a new 10% global tariff (may be raised to 15%) under a different legal authority. The court decision limited one mechanism, it did not end tariff policy.

2. Refunds are Delayed, not Immediate

Even if companies ultimately recover duties, that cash may not return for 12-18 months or longer. Retailers must operate on current cost structures, not future reimbursements.

3. Prices Rarely Reverse Quickly

Businesses that raised prices to offset duties are unlikely to reduce them immediately, especially amid continued uncertainty about future trade policy.

4. Amazon’s Warning Reflects Margin Reality

Before the Supreme Court struck down the IEEPA tariffs, Amazon CEO Andy Jassy said that the tariff “buffer period” was ending. He wasn’t talking about the legal details of one tariff policy, but he was focused on how long companies can afford to hold prices steady when their costs are rising.

In 2025, price impacts were modest because:

• Some tariffs were lower than initially announced.
• Businesses stockpiled inventory before increases.
• Companies absorbed costs to protect demand.

But much of retail operates on mid-single-digit margins. As Jassy put it, if costs rise 10%, there aren’t endless places to absorb that hit.

Economists at the Peterson Institute for International Economics warn that tariff pass-through to consumers may accelerate in early 2026 as pre-tariff inventories run out. Execution Over Uncertainty

The Supreme Court’s ruling clarifies tariff authority but doesn’t ease cost pressure, and Amazon’s CEO warns delayed price impacts are ending. In 2026, retailers that prioritize refund recovery, accurate landed cost modeling, and strong supply chain coordination will be better positioned to protect margins. Tariffs remain a constant variable, and data discipline will determine whether they erode profit or are managed effectively.

Other Amazon Sellers News This Week 

1. Amazon Expands Free, No-Box Returns

Amazon is expanding box-free, label-free “FREE Returns” to 8,000+ drop-off locations, making returns even easier for customers, and potentially more frequent for sellers.

As Amazon optimizes for speed and convenience, sellers may face tighter scrutiny around refund disputes and recovery claims. Now’s the time to review return rates, damaged inventory trends, and reimbursement processes. 

2. SAFE-T Claims: Recover What You’re Owed

As returns get easier, Amazon is urging sellers in the US and UK/EU to use SAFE-T claims to recover losses from buyer abuse, damaged returns, or incorrect refunds, with updated guidance on eligibility and documentation. If you’re not filing consistently or have faced denials, consider revisiting your process and strengthening your reimbursement strategy.

3. Tips to Keep Multiple Selling Accounts in Good Standing

Amazon is reinforcing its multiple-accounts policy: while separate accounts are allowed for legitimate needs, all related accounts must stay in good standing. A violation in one can trigger suspensions, listing removals, or payment holds across others. If you manage multiple storefronts, Amazon offers tips on how to protect your portfolio.

4. Amazon Ads Introduces AI-Powered Targeting

Amazon Ads has released new Product and In-market category targeting tactics across Display, Video, and Audio campaigns, powered by AI-driven multi-signal intelligence. These tools combine behavioral and contextual data automatically, reducing the need for multiple line items and manual targeting combinations. This could mean more precise reach with less complexity, if used strategically. 

5. Wayfair Returns to Annual Revenue Growth

Wayfair just reported its first annual revenue growth since the pandemic, with sales up 5.1% in 2025. US demand helped drive the rebound. For Amazon sellers, that’s a sign shoppers may be spending again on home and furnishings, which is something worth watching as we head into 2026.