For many businesses, the AI journey starts with efficiency: automating tasks, streamlining processes and reducing costs. This can deliver significant value, but it is unlikely to determine who ultimately wins the AI race.
The biggest competitive opportunities will come from using AI to create new products, services and business models built around intelligence, automation, personalization and better customer experiences.
Previous technology shifts followed the same pattern. Amazon reimagined retail around the internet, Netflix transformed entertainment through streaming and Uber rebuilt transport around mobile connectivity. AI is now creating a similar opening for companies prepared to rethink how they operate and where they create value.
The early leaders are already showing what this looks like. Here are five companies using AI to do things that previously were not possible.
Duolingo: Using AI To Build New Products
Duolingo is a great example of a platform that has used AI to create new products and services that it wasn’t able to offer before. These include opportunities for language learners to roleplay social and conversational scenarios with native speakers and to receive detailed and personalized feedback on their own progress. As one of the initial OpenAI partners, Duolingo has been repurposing its services as AI-first for several years now, resulting in strong revenue and subscriber growth. Generative AI is embedded across its business model, where it’s also used for generating and developing course content and learning pathways. In 2025, this let it double the number of language courses it offers in just one year, and reduced the time spent developing each one from several years down to one year.
Salesforce: Enabling The Agentic Workforce
Salesforce boss Marc Benioff has said that agentic AI, meaning AI that can take action, is so important to his company’s future that he might rename it Agentforce. This name is currently used for its agent-as-a-service platform, which is positioned as a gateway for every business into the world of AI virtual workers and automated assistants. Salesforce’s strategy is to become the “picks and shovels” seller for the AI gold rush, providing businesses with the tools they need to automate tasks and streamline workflows. Rather than simply using AI to solve its own problems, it is betting that providing others with access to an agentic workforce is the company’s future.
Klarna: Intelligent Customer Experience
Most people know Klarna as a provider of buy-now-pay-later financial services. But fewer are aware that it is also a major investor in AI, with a focus on creating more streamlined, friction-free online shopping experiences. As well as using AI extensively across its core credit application and decision-making functions, Klarna uses AI assistants, in partnership with OpenAI, to help shoppers discover new products, compare prices and make more informed purchases. Simultaneously, it has built AI out across its marketing and customer services operations to smooth customer journeys there. Klarna hopes AI will help it evolve from a simple payment processor to generating value at every touchpoint of the e-commerce customer journey.
Harvey: Democratizing Expertise
Harvey targets a niche audience, in this case lawyers, and provides AI-as-a-service tools for analyzing contracts, reviewing documents and drafting legal letters and submissions. The innovation here is that rather than focusing on routine administrative tasks, this is work that would generally be carried out by skilled professionals with expert understanding of what they are doing. This means human lawyers can take on more complex work, focusing their attention on human relationships and strategic outcomes. Harvey aims to solve problems faced by professional services companies due to the scarcity and expense of specialist skills and knowledge within the industry.
Adobe: Remaking Services Around AI
Adobe has long been one of the leading providers of creativity and productivity software. In its position, failing to adapt quickly to the arrival of genAI wasn’t an option, as more forward-looking competitors inevitably used it to build new tools and services. Adobe acted quickly to build AI functionality into its well-known packages like Photoshop and Illustrator. But it also designed innovative, AI-first services like Firefly, primarily sold via affordable monthly subscriptions and aimed at small businesses and freelancers. A key innovation is training its generative models only on data it has legally obtained the rights to, to give business users confidence they won’t run into legal problems in the future.
There are two common themes running through all of these success stories. The first is that they don’t involve AI primarily being used to cut costs. The second is that AI isn’t simply “bolted on” to existing services but used strategically to create new business opportunities.
Improving efficiency is critical, but only going for “low-hanging fruit” means opportunities to make real gains can be overlooked. My advice is to think about what AI can help you do that you could never do before, and start from there.
