Amazon vs Perplexity: The Legal Fight Over AI Shopping Agents
AI agents are moving from answering questions to taking actions, and one of the first big legal tests is now playing out between Amazon and Perplexity.
According to Reuters, a U.S. appeals court is looking at whether Perplexity can be held responsible for its agentic AI system allegedly accessing Amazon user accounts without Amazon’s permission. The case focuses on Perplexity’s Comet browser and shopping agent, which Amazon says could log into customer accounts and place orders on users’ behalf.
Perplexity’s argument is different. It says users are choosing to use an AI assistant to act for them online. Amazon’s argument is that the AI system is accessing its platform in a way it did not authorise.
That difference is important to notice because agentic commerce changes the normal rules of online shopping. If a customer logs in and buys something, the responsibility is fairly clear. But if an AI agent logs in, compares products, makes a recommendation and places the order, the question becomes more complicated.
For business owners, this is not just a legal story. It is a preview of where online commerce is heading. AI agents may soon compare suppliers, book services, buy products, renew subscriptions and make decisions before a human ever lands on a website.
The first-mover question is simple: if customers start using AI agents to shop, research and choose providers, will your business be ready to be found, trusted and selected by them?
The Briefing: What Else You Need to Know
Meta Is Turning Off-Platform Behaviour Into AI Personalisation
Meta says it will use activity from other websites and apps to personalise not only ads, but also Feed recommendations and Meta AI responses.
For brands, this pulls AI discovery deeper into the social web. The signals people create across websites, apps and platforms may now influence what they see in feeds and what Meta AI tells them. Search, social, and AI are starting to overlap.
EU Regulators Push Meta to Reopen WhatsApp to Rival AI Chatbots
EU regulators have ordered Meta to restore rival chatbot access to the WhatsApp Business API while an antitrust investigation continues.
WhatsApp is becoming more than a messaging app. It is already a sales, support and customer-service channel for many businesses. If AI assistants become a normal part of those conversations, the next platform battle may happen inside the inbox, not the app store.
Regulators Want AI Controls, Oversight and Kill Switches
Financial regulators are increasing scrutiny of how banks and financial firms use AI, especially as agentic systems become more common.
Reuters reported that the Financial Stability Board found 52% of financial-sector respondents were actively adopting agentic AI, with 23% scaling or transforming and 29% piloting these functions. AI governance is moving from policy documents to practical controls. The next enterprise question is not just whether AI can do the work. It is whether the business can control it when it does.
Anthropic Access Shutdown Turns AI Into a Geopolitical Risk
Anthropic reportedly disabled access to its most advanced models for some foreign users after a U.S. government directive, raising concern in Europe.
Businesses are now learning that AI access is not guaranteed infrastructure. A company can build critical workflows around one frontier model, then face a policy change, regional restriction or supplier decision that creates sudden operational risk. AI resilience may soon mean having multi-model, multi-region and sovereign AI options in place.
German Court Says Google Can Be Liable for False AI Overview Claims
A German court has ruled that Google can be directly liable for false claims generated inside AI Overviews.
AI-generated answers are no longer just search results. They are becoming statements that platforms may have to stand behind. For brands, publishers and reputation teams, this creates a new risk layer. If AI describes your company incorrectly, the damage may happen before anyone clicks through to your website.
Business Leaders Are Starting to Put AI “On a Diet”
Business Insider reported that some companies are introducing AI usage caps, switching models and demanding clearer return on investment as token costs rise.
The early AI phase was about experimentation and speed. The next phase is about discipline. When companies like Coinbase, Walmart and Amazon start controlling how employees use AI, it shows that AI is becoming a real operating cost. The question is no longer just “can we use AI?” It is “what is the return on each token?”
Visa Brings AI Agent Payments Into ChatGPT
Visa has integrated its payment network into ChatGPT, allowing AI agents to shop and complete purchases for users with permission.
This is a major step toward agentic commerce. If an AI assistant can recommend a product, compare options and complete payment, the customer journey starts to look very different. The future funnel may become ask, compare, approve and pay, all inside an AI assistant before a buyer ever visits a website.
Anthropic Puts $200 Million Behind AI’s Economic Impact
Anthropic has announced a $200 million investment to study AI’s economic impact, especially around jobs and labour-market disruption.
The companies building automation are now funding research into what automation could break. For founders and employers, the bigger question is not only how AI improves productivity. It is how teams, roles, training and trust need to change as more work becomes automated, assisted or redesigned by AI.
The Shift - How to Track AI Visibility
Most businesses know how to track Google rankings, website traffic and leads. Very few know how to track whether AI can actually find, understand and recommend them.
AI visibility starts with the questions your buyers are already asking. Not keywords. Real questions.
What would a buyer ask before choosing a company like yours? They might ask who the best providers are, which firms are trusted, how two competitors compare, what common mistakes to avoid, or which option is right for their situation.
Once those prompts are clear, businesses need to track three things.
→ First, are you mentioned? This shows whether the AI system knows your business exists.
→ Second, are you cited? This shows whether your website, content or third-party mentions are being used as supporting evidence.
→ Third, are you linked? This shows whether the answer can actually send people back to your business.
Those three signals are not the same. A brand can be mentioned without being cited. It can be cited without being recommended. It can even be described inaccurately because the AI is using old, weak or unclear information.
The practical question is not just, “Do we appear?”
It is, “When AI compares us with the market, are we being described clearly, fairly and competitively?”
This also changes the type of content businesses need. AI engines do not want vague sales copy. They need clear explanations, direct answers, comparison pages, evidence, pricing context, case studies, FAQs and structured information they can easily extract and trust.
The businesses that win this shift will not only publish more content. They will make their expertise easier for AI to understand, verify and recommend.
The Big Question for Businesses
For years, businesses have tried to control the customer journey through their own channels.
They built the website, optimised the funnel, ran the ads, improved the checkout, and trained the sales team. But AI is starting to sit between the customer and the business in ways that are harder to control.
It can shape what someone sees in a feed, answer their buying questions, compare providers, operate inside messaging apps, and move closer to completing purchases on their behalf. That creates a much bigger question than whether a business is using AI internally.
The real question is: who controls the customer relationship when AI becomes the interface?
If a buyer asks an AI assistant what to buy, which businesses get recommended first?
If an agent compares your business against competitors, what evidence does it use to make that comparison?
If a customer never lands on your website, how do you influence the decision they are about to make?
This is where businesses need to think beyond content, campaigns, websites and paid media. Your product information, pricing, proof, reviews, service pages, policies and reputation all need to be clear enough for AI systems to understand and trust.
Because the next customer journey may not start on your homepage anymore. It may start inside an AI-generated answer before your brand even knows it was considered.
What We’re Working On
1. The GATE Project: The Next Cohort Starts on 1 July
Our Founder Dean is currently in Cape Verde for The GATE Project, meeting officials, partners and candidates for the next intern cohort, which starts on 1 July.
The project is about creating access to the AI economy for talented people who may not otherwise get the opportunity. The next cohort will be trained in practical skills that businesses need now, including AI automation, agentic AI building and generative engine optimisation.
We are also creating a documentary around the project, showing the people, partnerships and real-world impact behind the work.
For businesses, there is a practical opportunity here too. If you want to request an intern from the next cohort to help automate your business, support AI workflows or bring trained AI capability into your team, placements are available from £600 per month.
To discuss The GATE Project or request an intern, email [email protected] or call 0121 517 2258.
2. Startup to Standup: Building in Public With a Real Business
We are also working on the Startup to Standup partnership, which will become a live, interactive, report-as-we-go documentary. The idea is to document how we do the work in real time, not just talk about strategy after the results have happened.
That means showing the decisions, systems, campaigns, AI workflows, content, commercial thinking and lessons involved in working with a real business and its members.
For founders and business owners, this should become a useful behind-the-scenes look at how strategy, marketing, AI and execution come together in practice.
You can watch the project unfold on the Build With Dean YouTube channel.
3. Answer Architect: What Is AI Really Saying About Your Brand?
Answer Architect can already scan your website and show how visible your business is in AI search. But the deeper question is not just whether your website can be scanned. The real question is what AI says about your business when buyers are actively comparing options.
That is where our AI Visibility Audit goes further. We look at the real questions your buyers might ask, which companies are being surfaced, which competitors appear instead of you, what sources AI is using, and how your brand is described in the comparison layer.
This matters because AI may not describe your business the way you would describe it. It may use old information, weak sources, missing content or competitor-led comparisons to form an answer.
A proper audit looks at how AI represents your business when someone is searching for a provider, but also how it compares you against competitors on things like price, quality, trust, reputation, service and proof.
That is where revenue can be won or lost.
If you want to know what AI is saying about your business, reply “audit” to this newsletter or email [email protected] to request a free AI Visibility Audit.
Final Takeaway
AI is no longer just answering business questions, the next phase becoming about control, visibility, trust and action.
Agents are moving closer to shopping and payments. Platforms are personalising what people see and what AI tells them. Regulators are asking where the guardrails, oversight and kill switches are. Businesses are starting to ask whether AI is delivering enough return to justify the cost.
The pattern is clear. AI is moving into the layer where customers discover, compare, decide and act. For businesses, the opportunity is not only to use AI internally. It is to make sure AI systems can understand your business, represent it correctly and recommend it when buyers are making decisions.
The risk is not that AI replaces your website overnight. The risk is that your next buyer gets an answer, comparison or recommendation before they ever reach it.
Want to Know What AI Is Saying About Your Business?
If you want to understand how visible your business is in AI search and how AI compares you against your competitors, we can help.
Reply “audit” to this newsletter, or email [email protected], and our team will run a AI Visibility Audit so you know exactly what AI is saying about your business.
Is this newsletter right for you?
We send this briefing to help first movers stay close to the AI news, trends, and signals that matter.
If that is useful to you, we would love to keep sharing it with you. If it is not quite what you are looking for, no problem at all, you can unsubscribe using the link below.
Either way, thank you for giving it a read.
