The AI Daily Brief: Artificial Intelligence News and Analysis - The Era of Agentic Shopping
Episode Date: October 3, 2025OpenAI has introduced Instant Checkout inside ChatGPT, allowing purchases to happen directly in the conversation. The update challenges the traditional search-to-cart model, raising questions about co...nsumer adoption, advertising, and control of online shopping intent. Discussion also covers the agentic commerce protocols powering this shift, Stripe and Shopify’s involvement, Google’s competing AP2 standard, and the broader implications for Amazon, Google, and the future of e-commerce.Brought to you by:Is your enterprise ready for the future of agentic AI?Visit AGNTCY.orgVisit Outshift Internet of AgentsTry Notion AI today with Notion 3.0 https://ntn.so/nlwKPMG – Discover how AI is transforming possibility into reality. Tune into the new KPMG 'You Can with AI' podcast and unlock insights that will inform smarter decisions inside your enterprise. Listen now and start shaping your future with every episode. https://www.kpmg.us/AIpodcastsBlitzy.com - Go to https://blitzy.com/ to build enterprise software in days, not months Insightwise - AI for the entire consulting lifecycle https://www.insightwise.ai/Robots & Pencils - Cloud-native AI solutions that power results https://robotsandpencils.com/Vanta - Simplify compliance - https://vanta.com/nlwThe Agent Readiness Audit from Superintelligent - Go to https://besuper.ai/ to request your company's agent readiness score.The AI Daily Brief helps you understand the most important news and discussions in AI. Subscribe to the podcast version of The AI Daily Brief wherever you listen: https://pod.link/1680633614Interested in sponsoring the show? nlw@aidailybrief.ai
Transcript
Discussion (0)
Today on the AI Daily Brief, buying stuff in chat, GPT and what it says about the business model of the internet.
And before that in the headlines, vibe coding comes to the enterprise.
The AI Daily Brief is a daily podcast and video about the most important news and discussions in AI.
All right, friends, quick announcements before we dive in.
First of all, thank you to today's sponsors, pencils and robots, Notion, KPMG, and Insightwise.
To get an ad-free version of the show, go to patreon.com slash AI Daily Brief,
or you can now subscribe directly on Apple Podcasts, Spotify should be coming within the next few days.
Finally, if you're interested in learning about sponsorship opportunities or speaking engagements
or basically anything else about the show, go to AIdailybrief.aI.
There are a bunch of email links there as well.
With that, let's dive in.
Welcome back to the AI Daily Brief Headlines edition, all the daily AI news you need in around five minutes.
We kick off today with Salesforce bringing vibe coding to the enterprise with a new AI developer tool.
Agent Force Vibs is a new AI-powered IDEE designed to build, test, and deploy Salesforce apps, and agents.
The platform is compatible with VSCode IDs like Cursor and WinSurf and leverages Klein for agendic chat functions.
In addition, the platform includes an AI coding agent called VibeCode.
Gotta anthropomorphize these things.
It's basically a rule of nature right now.
The agent is designed to work as a pair programmer generating and refining code based on text prompts.
As with other Salesforce AI products, the platform is built on.
OpenAI's models, in this case, GPT5.
Now, of course, the big selling point for this is having a vibe coding platform that's
already plugged into your organization's Salesforce environment, allowing it to reuse
existing code and follow coding guidelines.
Dan Fernandez, the VP of Product for Developer Services said, we're trying to give you
everything, so rather than having to spend a bunch of time on setting up MCP, setting up
a dev environment, setting up tools, everything's pre-built and ready for you, including AI
request to get started, and that reality is sort of a differentiator on how we're lowering
the barrier to entry.
A couple different takes I've seen on this.
The first is, yes, a bunch of people are noticing the names.
V.C. Tenai Jopora writes, absolutely incredible that Salesforce is really calling this
agent force vibes and the agent within it vibe Cody.
I also saw more than one person comparing it to Microsoft's classic old clipy.
There were also some negative takes that basically argued that this is Salesforce, throwing
anything with agents at a wall to see what sticks.
But my instinct is that it's not that at all.
I think instead what this is validation of is the fact.
that in eight short months since this term was coined,
vibe coding and the broader agendic coding that it sits within,
has gone from something that was nascent,
just for individuals,
and often finding resistance in the enterprise,
to something that is completely transforming businesses from within.
Agendic coding has become the most dominant use case for AI inside companies,
but what it takes to do agentic coding inside companies
is different than what it takes to build a great consumer tool.
The market is absolutely starting to fill in that gap,
Factory, for example, is a much-beloved company that is explicitly going after that,
and I think that this is just an absolutely enormous growth area, and one that it makes
complete sense for Salesforce to try to get in on.
Now, staying on the agent decoding theme for a minute, Google's Jules' coding agent
is now available in Ide's via a public API as well as command line interface.
Previously, Google's asynchronous coding agent was only available through a web interface and GitHub.
Now developers can summon Jules within their normal IDE, which of course eliminates costly context
switching. Now, Jules is doing a lot of work to be a player in the AI coding game. They have been pumping
out features and continue to tease new ones, just like I was just saying that it made sense to me
that Salesforce, as an existing big B2B enterprise-style player is going to play in this vibe coding space,
so too it makes sense that Google is putting a ton of emphasis in this particular product.
Staying on our enterprise theme for a moment, Claude has entered the chat literally with Anthropics Chatbot
now available as a native plugin for Slack. Users can now call on and tag Claude as a
they would with any coworker. Within Slack, Claude can search through channels, DMs, and files for
additional context. Anthropics suggests prepping for meetings, gathering project updates, or creating
documentation, all of that juicy, full Slack context. Now, some noted that this is actually
the second time Anthropic has announced Slack integration with the previous launch all the way back
in early 2023. That iteration of Cloud for Slack was made impossible in June when Salesforce
changed Slack's API policy to prevent third-party AI products from accessing chat data. That
change severely impacted AI products built on top of the corporate data contained in Slack, particularly
Enterprise AI search engine Glean. It also highlighted that Salesforce recognized that the rich context
available within Slack was a valuable resource in the AI era. From content is king to context
is king. Alongside this week's announcement from Anthropics, Salesforce announced the general
policy change to allow third-party AI apps back into their ecosystem. On Tuesday, Salesforce announced
the new API for real-time search for Slack, an MCP server, and in
hence developer tools for AI developers.
There was a long list of partners who are already building AI apps and agents designed to
plug into Slack, however, Glein was not on that list, with Salesforce seemingly wanting to
control the context and search layer themselves.
They wrote, conversational data is the gold of the agentic era, yet it's been locked away in
unstructured messages and chats, largely out of reach for employees, let alone applications.
TLDR is that Salesforce appears to be happy to make Slack a multi-agent interoperable work platform,
while also recognizing the need to control that digital gold of the context layer.
One more on Anthropic, the company has hired a new CTO with a focus on infrastructure.
Former Stripe CTO, Rahul Patil, will join the company taking over from co-founder, Sam McCandlish.
McCandlish will move over to a new role as chief architect.
As part of the change, Anthropic will also restructure their teams, bringing the product
engineering team closer to the infrastructure and inference teams.
Patil will oversee compute infrastructure inference and assorted engineering tasks,
while McCannlish will be responsible for model training.
This shifting focus to infrastructure makes a ton of sense given industry trends
and recent Anthropics-specific challenges.
On the one hand, you had those high-profile infrastructure failures over the summer
that really undermine Anthropics' reputation for a while there,
but more broadly, you just have the incredible competitive pressures that they're operating within.
OpenAI is over here announcing multi-100 billion-dollar infrastructure deals,
but we've heard very little about Anthropics infrastructure plans.
Are they still largely reliant on their strategic partnership with AMN?
Amazon to secure the compute they need.
Those are the sort of questions that are going to get louder and louder.
One update from the browser wars,
Perplexity has made their Comet AI browser available for free.
Until now, Comet was exclusively available to Perplexity Max customers,
i.e. those who are willing to pay $200 a month to cover the high cost
of serving an agendic internet experience.
Free users will have a very similar experience to the original release.
They'll be able to access the in-browser assistant, shopping, travel, and finance tools.
However, the free browser will use less powerful AI models, which could limit some of its agentic
features.
Max subscribers will still have access to higher-performing models alongside Perplexity's
newly released email assistant.
Now, for many, it is tempting to count Perplexity out a priori because of just the incredible
distribution of their competitors, but having a free version does really make a big difference.
Lastly, today, OpenAI has closed their monster secondary round, officially becoming the most
valuable private company in history.
The deal closed at a valuation of $500 billion, pushing the new.
them past SpaceX at $400 billion for the first time.
Current and former employees sold around $6.6 billion worth of OpenAI stock to investors
including Thrive Capital, SoftBank, Dragonere, Abu Dhabi's MGX, and Tiro Price.
Now, OpenAI had allowed for up to $10 billion to be sold, so it seems that a lot
of employees are holding onto their stock rather than grabbing the liquidity.
Many residents of San Francisco lamented the fact that that $6.6 billion is more than the
entire amount that was spent on houses in San Francisco last year, meaning basically that
if you were not at OpenAI, you don't get to buy a house in San Francisco for the foreseeable
future. In one hilarious twist, Elon Musk's net worth also just recently tipped over $500 billion,
with Lassano Give joking, Open AI is now worth one Elon Musk. We don't have nearly enough time to get
into that can of worms. Indeed, for now, that is going to do it for today's headlines.
Next up, the main episode. Small, nimble teams beat bloated consulting every time.
Robots and Pencils partners with organizations on intelligent cloud-native systems powered by
AI. They cover human needs, design AI solutions, and cut-through complexity to deliver meaningful
impact without the layers of bureaucracy. As an AWS-certified partner, Robots and Pencils
combines the reach of a large firm with the focus of a trusted partner. With teams across the
U.S., Canada, Europe, and Latin America, clients gain local expertise and global scale. As AI evolves,
they ensure you keep peace with change. And that means faster results, measurable outcomes,
and a partnership built to last.
The right partner makes progress inevitable.
Partner with robots and pencils at robots and pencils.com
slash AI Daily Brief.
Chatbots are great, but they can only take you so far.
I've recently been testing Notion's new AI agents,
and they are a very different type of experience.
These are agents that actually complete entire workflows
for you in your style,
and best of all, they work in a channel
that you already know and love
because they are purpose-built Notion super users.
Notion's new AI agents completely expands the range of what Notion can do. It can now build documents
from your entire company's knowledge base, organize scattered information into organized reports,
basically do tasks that used to take days and get them complete in minutes. These agents don't just
help with work, they finish it. Getting started with building on Notion is easier than ever.
Notion agents are now your very own super user to help you onboard in minutes. Your AI teammates
are ready to work. Try Notion AI for free at the link in our show notes.
What if AI wasn't just a buzzword, but a business imperative?
On You Can with AI, we take you inside the boardrooms and strategy sessions of the world's most forward-thinking enterprises.
Hosted by me, Nathaniel Wittamore, and powered by KPMG,
the seven-part series delivers real-world insights from leaders who are scaling AI with purpose,
from aligning culture and leadership to building trust, data readiness, and deploying AI agents.
Whether you're a C-suite executive, strategist, or innovator, this podcast is your front-row seat,
to the future of Enterprise AI.
So go check it out at www.kpmg.org.us slash AI podcasts or search you can with AI on Spotify,
Apple Podcasts, or wherever you get your podcasts.
As a consultant, responding to proposals can often feel like playing tennis against a wall.
You're serving against yourself trying to guess what the client really wants.
That all changes with the Insight Wise Proposals platform.
Now you've got an AI coach that thinks just like your client.
It returns to the brief time and time again, identifying opportunities, showcasing your track record,
and making recommendations to improve your pitch.
Suddenly, you're on center court, but this time you've got a secret weapon.
Insightwise does away with all the time-consuming manual work so you can focus on winning more business more often.
Generate reports, pull insights from your own data, build competitive advantage and go to sleep before 2am.
When it comes to proposals, you only get one shot.
With Insightwise, make yours an ace.
Go to insightwise.aI to start your free trial today.
Welcome back to the AI Daily Brief.
It has just been an absolute barrage of updates and new announcements this week.
And frankly, I don't think we're slowing down anytime soon.
But today we're finally catching up on an interesting announcement from Monday.
ChatGPT has announced a new ability to buy things from within ChatGPT and an agentic commerce protocol to support it.
Now, I think it's important to put this in the context of the broader shift in the business model of the internet that AI is generating.
For a very long time, the core business model of the internet has been the relationship between
search and websites where you could either consume stuff and get served ads or buy stuff and
give them your money. Basically, the way that people entered the world of the internet was via
search. And from search, they would be directed to websites, which depending on the nature of those
websites, would either sell them stuff or serve them ads. And even as social started to become
a force and people were spending more time in those channels, that relationship was still largely
intact. It just shifted around the players a little bit. AI represents a fundamentally different
consumption pattern for information on the internet. More and more, people are opting to use
some version of chat GPT or even in Google AI overviews to get the information they want quickly,
rather than having to click around to a bunch of websites to get it. Now, there is a lot of debate right
now around just how dramatic the impact is. Surveys and studies are telling slightly different stories,
and certainly the AI companies are trying to say that this is all value accretive, but it does
seem like for some portion of people, AI summarization, be it in the traditional channels like
the AI overviews of Google or the new channels like ChatGBTGBT, is replacing those old patterns.
It's enough of an issue that some in the industry are coming up with their own answers.
Cloudflare, for example, is now blocking AI web scrapers by default and trying to create
a whole new system whereby AI companies can explicitly pay companies to access their content.
And yet, if the destruction part of this creative destruction is clear, where less clicking around
Google Links means less ultimate payment and revenue for publishers, the creative part is also
starting to come into focus. AI traffic, as you might imagine, is growing in spectacular fashion.
A study from Previsible found a 527% increase in AI referred sessions between January and May 2025.
You might have seen SEO people talking about generative engine optimization or geo instead of
SEO. And so even as one type of pattern changes, another is being born and that new pattern
creates opportunity as well.
And that is the context into which these new agentic shopping features come.
In announcing their new instant checkout and the Agent Commerce Protocol, OpenAI wrote,
US users can now buy directly from US Etsy sellers right in chat,
with over a million merchants like Glossier, skims, Spanx, and Vori coming soon.
With a big bold pronouncement, they say, this marks the next step in agentic commerce,
where chat GPT doesn't just help you find what to buy, it also helps you buy it.
For shoppers, it's seamless.
Go from chat to checkout in just a few taps.
For sellers, it's a new way to reach hundreds of millions of people while keeping full control of their payments, systems, and customer relationships.
Right now, the feature only supports single item purchases, but they say multi-item carts as well as expanded merchants and regions are coming soon.
Fortune gives an example of how this might work in practice.
They write, when a chat GPT user asks a shopping-related question, such as the best hiking boots for me that cost under $150 or possible birthday gifts for my 10-year-old nephew,
the chat bot will respond with product suggestions.
Under the new system, if a user likes one of the suggestions and instant checkout is enabled,
they'll be able to click a buy button in the chat bot response and confirm their order,
shipping and payment details without ever leaving the chat.
Now, as you might imagine, people immediately jumps to the idea that this was going to be another
vehicle for advertising, considering that the SORA app content feed and ChatchipT Pulse's
hyper-personalized newsletter thing came out last week, there are a lot of people looking over
at OpenAI saying ads have to be on their radar in a big way.
Still, at least for now, OpenAI said that their product results are, quote, organic and unspensored,
ranked purely on relevance to the user.
They also say that merchants can't pay additional fees to be ranked higher.
Because Chatchapit subscribers who already pay have their credit card information already there,
they can use that same card or store an alternative payment method.
Now, this is definitely rolling out very slowly.
In fact, I've seen almost no one talk about the actual experience.
One that I was able to find came from Glenn Gabe, who wrote,
I just triggered a product from Etsy with instant checkout possible in ChatGBT.
It was pretty slick and very easy to buy.
Uses your credit card on file, but you can pay via another source.
I didn't complete the purchase since I was just testing it out,
but this could be a big way for OpenAI to drive affiliate revenue.
That's if people are comfortable buying this way without visiting the site and reviewing everything there.
Note you can visit the site via ChatGPT, there were several links to see the reviews,
visit the product page, etc.
Now, alongside the new feature, OpenAI discussed the infrastructure that underpins it.
Stripe is the major partner for the feature and the technology, which consists of two main elements.
First, there is the agentic commerce protocol or ACP.
The protocol sets a standard format for communications between buyers and merchants with agents in the middle.
The idea, of course, being to facilitate fully programmatic commerce.
They write, when someone places an order, chat GPT sends the necessary details to the merchant's
back end using agentic commerce protocol.
The merchant accepts or declines the order, processes the payment via their existing provider,
and handles fulfillment and customer support exactly as they do today.
ACP allows merchants to define the key features of each item as machine-readable code,
and on the purchaser side, ACP allows their agents to define all the relevant details of an order.
ACP is being released as a fully open standard so it can be adopted by any merchant or payment provider,
and if a merchant already uses striped process payments, they can enable ACP with just a single line of code.
Now the other half of the system is a new API for agentic payments called shared payments tokens.
This is the mechanism that allows an agent to transmit payment details, including verifying usage limits,
as well as expiration windows. Shared payment tokens are also compatible with any payment services
provider. Shopify continues to be on the cutting edge with CEO Toby Lucki, saying that they've had a hard
time keeping this announcement quiet. Stripes Patrick Collison reinforces that there is a big change
afoot. At the end of his announcement tweet, he said, it's clear that internet purchasing
modalities are going to change a lot and we're excited to start to lay some of the foundations.
Vayner Commerce President Zubin Malawi said,
Google and Amazon have been the toll booths of online shopping for 20 years. OpenAI's new
buy it and chat GPT signals a different future. This isn't about checkout. It's about who controls
consumer intent. For two decades, he continues, Google sold clicks off intent, Amazon sold shelf
space off intent. Control discovery, control margins. That was the model. ChatGPT collapses it.
Search, click, click, cart, checkout becomes one conversation. That rewrites the playbook. Instead of
ranking on Google's page one or winning Amazon's buybox, brands must rank in the AI's answer set.
Basically, he's saying there is a new gatekeeper in town and that gatekeeper.
is ChatGBTGBT. Bold Metrics co-founder Morgan Linton writes,
I'm too excited about agendic shopping. I think it's a foundational change that's going to
completely transform how all of us shop. So much so, in fact, that he started a new
substack just for this. In the first post, he writes, this week OpenAI made a big
announcement. One, I think, will make this week the tipping point for a foundational change
in how consumers buy everything. I know that sounds dramatic, and while it is, like most
change, it won't happen overnight. Consumer habit changes in particular take time. Morgan
argues that this represents two big shifts for consumers. The first is that consumers now basically
all have an agentic personal shopping assistant, and the second is that they never have to visit
the retailer's website. Morgan basically thinks that the previous experience of browsing products
in a catalog like interface, and then ultimately entering credit card information, is ultimately going
to feel draconian and old. Cal Short thinks that there are some big market implications. He says,
saw this two months ago and chucked 100K into shop, which is of course the Shopify stock. He continues,
I think there's a few big movements this creates.
One, AI is the new marketplace, Amazon loses clicks.
Two, Google will be forced to respond with G-shopping.
Three, Shopify revenue from merchant benefits long term.
Nathan Lambert, like Morgan, thinks that this is a big deal.
In fact, he thinks it's a bigger deal than Claude 4.5 Sonnet, which came out on the same day.
He also wrote about this on his blog.
In that post, he writes,
the questions have long been whether AI models actually have the precision to find the items you want,
to learn exactly what you love, and to navigate the web to handle all the corner cases of
checkouts. But he says, I'm cautiously optimistic about this. Finding great stuff to buy on the web
is as hard as it's ever been. Users are faced with the gamification of Google search for shopping
and the insh-sha-facation of the physical goods crowding out Amazon. Many of the best items to buy
are found through services like meta's targeted ads, but the cost of getting what you want should
not be borne through force distraction. Open AI will not be immune to the forces that drove
these companies to imperfect offerings, but they'll come at them with a fresh perspective.
perspective on recurring issues and technology. If this works for OpenAI, they have no competitor.
They have a distribution network of nearly a billion weekly users and no peer company ready
to serve agentic models at this scale. Yes, Google can change its search feed, but the
thoroughness of models like GPT5 thinking is on a totally different level than Google search.
This agentic model is set up to make ChatGBT the one agentic app across all domains.
Now, certainly the market liked it. On the announcement, Shopify was up 6% and Etsy was up 16%.
And so the question really does become, is this just a normal.
part of the chat GPT experience. On the one hand, it feels like you're going to see a ton of grumbling
and scuttlebutt as you always do with new features, where people who have anything that whiffs
of an ad or an attempt to part them from more dollars in their wallet as something in annoying
and infringing on the pure experience that they want out of their chatbot. But for others,
it's going to seem extremely convenient. It's going to allow them to, to Nathan's point, not have
to click around endlessly to find the things that they're looking for, to be able to power
more of their lives from within this core experience that they're already using, and basically just
have a simpler, more fluid shopping experience. Now, my strong instinct is that this doesn't change
everything all at once. It's important to remember when we think about the potential impact that
the catalog style browsing isn't just a concession to UI, it's also an experience in and of itself.
A cheesy but very real example, you better believe that around August of every year I check out
what's new in Grandin Rhone's Halloween Haven, and this is not an experience that I want to short circuit.
I want to go see every single thing they have, new and old.
And look, now that we're in October, there's some crazy savings.
I am definitely going to get the Fountain Immortal Statue for our front yard.
But the point is that there is more to shopping than just efficiency,
and I believe that the very valuable efficiency of Buy It in Chatchipit will become a part of the overall
commerce tapestry.
Now, one little interesting sub-story is that ACP is actually the second agentic shopping protocol
released over the past month.
A few weeks ago, you might remember that Google announced their own.
version called AP2. AP2 was launched with Coinbase as the tech partner and over 60 merchant and payment
partners. That list included Visa, MasterCard, PayPal, Union Pay, World Pay, and Revolut, so a wide
cross-section of global providers. Shopify wasn't mentioned in the press release, but Etsy was also
on board with that one. Google explained a lot of the design considerations when it came to
agentic shopping. They noted that there are three big problems that AP2 seeks to solve. Authorization,
authenticity, and accountability. For authorization, agents need a way to prove
that a specific user gave it authority to make a purchase. For authenticity, agents need to be
able to prove that an order matches the authorization they've been given. AP2 allows orders to be
defined both strictly and generally. For example, a user can tell the agent to buy a specific
pair of shoes for a specific price, or they can define a budget and ask for a general type of
purchase such as a pair of jeans. In either case, the agent can communicate this intent to the
merchant system to prove the request matches the user's instructions. Finally, accountability
deals with acknowledging where the liabilities lie if there is a fraudulent or incorrect
transaction. Human users could always claim that they didn't approve the purchase and claim a chargeback,
making the entire concept of agentic commerce fail. Now, what's interesting about this to me is that in
almost every other area of agents so far, companies have seemed to behave as though it mattered more
to move together on common standards than it did to try to own the standards. And yet, from where
I'm sitting, AP2 and ACP look pretty directly competitive. Format wars when it comes to these types of
protocols are not really great for consumers. They slow things down, they make things more annoying,
and so I will be keeping an eye to see on how that engages. But it is certainly the case that the fact
that there are two agentic payments protocols released in a very short period of time is further
evidence that this change is coming to shopping, whether we want it to or not. So buckle up,
get ready, and start buying stuff in chat GPT. That's going to do for today's episode. Appreciate you
listening or watching, as always. And until next time, peace.
