In certain instances, we may earn revenue when a reader makes a purchase or completes a form through our pages or on a partner’s website. We adhere to a strict 
The US Department of Justice’s antitrust case against Google has prompted buzz that OpenAI is eyeing the Chrome browser. An OpenAI executive testified that the company would be interested in buying Chrome if the court forces Alphabet to sell the business, and that Google previously rejected a bid to use its search technology within ChatGPT.
By controlling both Chrome and its AI models, OpenAI could streamline real-time AI interactions. At the same time, OpenAI is challenging Google in commerce, introducing product recommendations and buy buttons directly within ChatGPT.
If OpenAI controls both the endpoint and backend, it’s a play for vertical integration across model, processing, and user interface that could shake up the industry.
Techopedia spoke with Darren Kimura, CEO and president of AI Squared, about what this would mean for small and mid-sized enterprises.
Key Takeaways
- The possibility of Google Chrome being potentially spun out into an independent company and potentially acquired by OpenAI would upend online search and commerce.
- The vertical integration of OpenAI potentially presents a new antitrust risk similar to Google.
- Vertical software-as-a-service (SaaS) businesses, as we know them, could soon be obsolete as AI becomes more embedded into our lifestyle and learns more about us.
- There are new opportunities for AI-native companies that are building agentic orchestration to create seamless, efficient applications.
- The human must be in the loop for the foreseeable future to override and log where AI mistakes are made, so that the models can improve.
About Darren Kimura
Darren Kimura is the CEO and President of AI Squared. Darren Kimura brings over 25 years of experience in scaling technology companies and bridging AI with enterprise solutions. Prior to his recent appointment, Kimura joined AI Squared in 2024 as President and COO and has been instrumental in shaping the company’s operational growth and strategic direction.
Kimura also served as President and CEO of Energy Industries Corporation, Solar Power Technology, and LiveAction.
The Potential Impact of OpenAI Moving into Chrome
A: The possibility of Chrome being potentially spun out into its own company, and having OpenAI invest and embed technology into it, would effectively displace a lot of what Google is doing. Google search, for example, could be removed or replaced. You would be using the OpenAI models and not the Gemini models, so it changes everything.
It could happen in the form of a browser extension, for example, and it would give OpenAI access to over 3 billion users. OpenAI is already doing incredibly well with its new raise at a $100 billion valuation. But this would give it access to everything else, including education, which is already so Chrome-dependent and an area OpenAI is trying to get into. It would be a total game changer.
Open AI does have a broad reach, but it’s not a household name yet. People are beginning to use it – and it has 40 million users per week – but this will take it to the next level, globally.
Chrome is everywhere in the world. Chrome in its entirety is like an operating system, and it is embedded in everything; it’s web browsers, it’s the Chromium engine – even Microsoft uses it rebranded.
It is the power behind Chrome laptops. It could even be embedded in mobile devices, because Android devices rely on Chrome. So this is enormous.
A: Imagine a world where Chrome is now powered by OpenAI. The large language model (LLM) will be learning everything about you – where you’re going, what websites you’re visiting, and understanding your patterns and behaviors.
And what I envision next is an agentic workflow where those agents would be acting on your behalf; they would automatically spin up agents to interact with websites based upon your past behavior.
If you go to a web-based application, rather than having you input or sort data, it will know what to do. It will prompt you to find out what you want to do and make it into one automated process.
You’ll be able to get to your data much faster. And that’s just an example. It could be in anything, and it could do it all automatically and agentically. It will give you time back. It will be a lot smarter, and it can be more accurate over time. This will change everything about the way we work.
Competitive Risk or Capitalism at Its Finest
A: It does. What we’re seeing with Google is just the beginning. OpenAI is arguably bigger than the majority of publicly traded companies already. It is reportedly at a $100 billion valuation, and that could double in a month or two. There are going to be massive companies like it and others that will need regulation as well.
There is no way around that – it’s capitalism at its finest. There will have to be regulatory oversight over a company that big. At one point, it seemed Google would never be displaced, and it would become the next biggest company in the world. Now it is under threat of antitrust, and you could argue that it slipped at the beginning of the AI journey, while OpenAI is taking over.
One of the big differences is that Google is heavily driven by ads, whereas OpenAI is finding revenue through efficiencies.
The technology provides such a stark difference in efficiency and accuracy that users are willing to pay for a subscription, unlike in the past when search had to be free. The ad model is changing.
A: This may be a bit provocative, but the future of verticalization is over. Consider companies that provide solutions, like Stripe, Block, WooCommerce, or Shopify, which have built their entire existence around online shopping.
As AI becomes more embedded into our lifestyle and learns more about us, we don’t need to have those vertically specific types of companies anymore.
An AI model could effectively replace what a company can do.
From an end-user perspective, the models can procure or curate content for you through the browser. Your browser becomes your interface, your activities and behavior patterns become what’s driving it, and behind the scenes are these small language models (SLMs) that live within the browser itself.
These companies will probably go away. You’re going to have more focus on AI infrastructure-type companies and model building, as opposed to verticalized product-type companies.
A: It is. I think vertical SaaS is dead. Some of the companies are so big that they’ll figure out a way to innovate out of it.
Salesforce, being the biggest of them all, has been acquiring companies like Slack, Tableau, and other technologies to give it a moat. It is using those acquisitions and technologies to create barriers for competitors. But other companies don’t have that.
You will eventually be able to do all the same things you would do with a CRM by using a broad data source, like Databricks or Snowflake.
An AI infrastructure platform would allow you to pick a CRM model that will go into your data store, find all the relevant pieces of data, combine, curate, and serve it to you in a way that a CRM used to be able to do, but without having a CRM.
A: It will take a couple of years. Most of the AI models we’re seeing today are still going from POC to production.
Companies in the small-to-mid-sized market are probably going to be fine for five years, but at some point, the platforms are going to take over. They could be OpenAI, Perplexity, or who knows?
Opportunities for Innovation
A: Absolutely. There are companies out there today that are building agentic orchestration platforms that don’t care about the vertical.
You can build an agent with a couple of clicks and no code, and deploy an entire agent stack. In a couple of years, they’ll pick up all those use cases as they go, and they’ll have these recipes that you can select to deploy.
Say you want to create a marketing automation campaign. Pick a drop-down item, select your motto, pick the data sources, press run, and you’ve now got marketing automation covered.
You don’t need a human – sales or business development representatives – just one platform and model.
When you build AI-native, it makes everything a lot better and faster, because it is seamless. When you build on top of an old monolithic software stack, it is hard to make it more efficient and faster.
These new companies that are coming in a true AI-native way are far ahead.
A: The human has to be in the loop, at least for the foreseeable future, before the results go into the real world. There are still going to be mistakes.
The human must have the ability to override it, to log where the mistakes are happening, so that the models can effectively improve.
As we create more insights, they are fed back into the source and could influence the core of the model. You can have third-party bias detection, which allows you to understand if the model is biased, but the human is at the end.
How to Approach Ethics & Regulation
A: It is a pipeline where you have data sources, the models that you are running, and then the insights that are being produced.
At that point, you need to insert filters like third-party bias detection or fairness review. It could be redaction if there are certain pieces of information that should not be shared, for example, to comply with GDPR in Europe or CCPA in California.
It could be cybersecurity reviews to make sure there is no malware accidentally or intentionally planted that the human may not be able to catch.
You can have many different filters, so that by the time it gets to the human, it will have gone through a lineage of checks to make sure the result is good.
A: It’s definitely not where it needs to be.
We need more regulation in AI because it is inconsistent. NIST has a framework for AI, but it is moving so fast, week by week. It is difficult for regulation to follow that because you would need governance bodies and standards to be able to oversee and peer review developments over time.
Different countries take different perspectives on the use of AI, particularly about cybersecurity and red teaming, and need to align their approaches. There is a lot to do there.
A: It would be helpful and important for society if countries agree on cyber secure red lines for AI, to set out what is acceptable.
It is similar to other pacts that we’ve had, whether it be nuclear or climate. The most exposed area is cybersecurity, given the way these systems are learning and changing in real-time, and how horrible it could be if they were let to run amok.
These agents that we described for the productive benefit of mankind could also be used for destruction.
The Bottom Line
If OpenAI were to control Chrome, it could reshape the web itself, disrupting search, commerce, and vertical SaaS models.
While this opens vast new opportunities for AI-native businesses, it also raises fresh antitrust and regulatory questions as traditional software stacks could face extinction.
