Klarna CEO on How AI Is Changing Banks and Jobs

Key Concepts Digital Financial System, Customer Mobility, Balance Sheet Banks, Digital Financial System Providers, Fintech Disruption, Legacy Banks, Modern Tech Stacks, AI Bias, Hallucinations, Knowledge Work Shift, Revenue per Employee, Regulatory Stifling, Consolidation in Fintech/Banking, Cost of Software Production, Global Financial Technology Companies, AI-assisted Coding.

AI's Reshaping of the Banking Landscape

The speaker envisions a future where AI fundamentally transforms retail banking, a vision held for over a decade. This future entails a digital financial system that proactively analyzes customer spending, such as mortgage payments, to identify better offers. The AI would then handle all necessary paperwork and processes, requiring only a "yes" from the customer, thereby saving costs. This paradigm shift is likened to the inevitability of self-driving cars.

Currently, banks often lack competitive pressure because customers face significant hurdles in switching providers, including the difficulty of transferring data. However, AI will usher in an era of ultra-high customer mobility, where individuals will readily switch to providers offering the most value for the least money. This increased competition is expected to drive down excess profits within the retail banking industry.

Vulnerability of Legacy Banks and Future Business Models

Legacy banks are particularly vulnerable to AI disruption. The speaker suggests that many will likely evolve into "pure balance sheets," primarily focused on optimizing for return on equity. The core challenge, highlighted by an early digital assistant developed by a company acquired in 2011, is that such assistants enable payments without direct bank logins. Banks initially "hated it" because it made their brand irrelevant; the value provider (the assistant) became paramount, not the underlying bank (e.g., Barclays).

This leads to two distinct paths for banks:

1. Balance Sheet Provider: Focus on capital management, accepting that profits will be closer to a "perfect market" scenario with minimal excess.
2. Digital Financial System Provider: Strive to be the entity that provides direct customer value through AI-driven services. This path is more challenging and fewer will succeed, but it represents "where the future of value of banking really is."

Legacy Banks vs. Fintechs in the AI Era

The competitive landscape in the AI-driven banking sector will feature three main types of players: big tech companies (e.g., Google, Amazon), fintechs, and incumbent banks attempting to reinvent themselves. The speaker argues that companies like Klarna and Revolut, which started as fintechs and are now "fully regulated entities" with "modern tech stacks," possess a significant advantage. Unlike legacy banks burdened by outdated infrastructure like "Cobalt Code running in the back mainframe," these modern fintech-banks are unencumbered by such challenges.

While some incumbents, like Jamie (presumably Jamie Dimon) or David Solomon with Goldman Sachs' Marcus, have attempted reinvention, many banks halted their fintech aspirations when valuations dropped in 2021-2022. The speaker emphasizes that "FinTech never die"; the downturn was merely an "investor sentiment" issue, as fintech revenue and customer adoption continued to grow. The industry is now "waking up" to the ongoing disruption, evidenced by the resurgence of companies like Klarna and Affirm.

Addressing Trust and Bias in AI Banking

The issue of trust and potential bias in AI, particularly in sensitive areas like lending decisions, is acknowledged. The speaker notes that Klarna currently avoids using much AI or machine learning for "writing" (lending) due to these risks. However, AI and Large Language Models (LLMs) are "extremely efficient and helpful" in other domains.

The speaker expresses confidence that as the technology progresses, methods to avoid "hallucinations or the issues of trust" will be found. An example cited is Klarna's success with AI in dispute management. An LLM consistently processes evidence from merchants and consumers to resolve disputes, demonstrating "higher quality and more consistent" decision-making than humans, who can become "bored" or less consistent over time. This consistency can build trust, though it requires careful "engineering and some thoughtfulness" in implementation.

AI's Impact on the Financial Services Workforce

AI is predicted to cause a "massive shift" in knowledge work across all sectors, not just banking. The speaker highlights the example of 8,000 translators in Brussels, whose work could largely be automated by AI, even with human quality assurance. While new jobs will emerge, society faces the short-term challenge of supporting displaced workers (e.g., a translator won't instantly become a YouTube influencer).

Klarna's own experience illustrates this shift:

• Revenue per employee increased from $400,000 to over $1,000,000 in the last two years.
• The company's workforce shrunk from 7,400 to 3,000 people, while revenue and customer numbers grew significantly.
• Klarna avoided layoffs by "simply not recruiting" and reinvested payroll savings into "acceleration of the compensation of our employees," providing a "huge benefit" from internal AI adoption. The ultimate goal is to leverage these technologies to drive maximum value for customers, acknowledging the broader implications for knowledge-based work.

Regulatory Concerns and Market Structure

The speaker identifies a "huge risk" in Europe regarding AI regulation potentially stifling innovation, though the regulatory direction has recently "shifted a little bit." A key concern is that non-democratic countries are aggressively advancing AI, and democratic nations should strive to remain competitive rather than impede progress.

AI will also create "true customer mobility" in banking, a phenomenon largely absent for a long time. This means funds and deposits will flow more freely between financial institutions, posing new challenges for banks, such as managing liquidity risks if customers rapidly move their balances. Overall, this increased mobility is seen as beneficial for society, leading to "less excess profits" in markets that were previously not optimally functioning.

Regarding market structure, AI is expected to lead to consolidation rather than more entrants. The "cost of producing software is going towards zero" because machines can now generate code that was once painstakingly written by highly paid engineers. As the cost of software manufacturing rapidly declines, "excess profits in software technology will be less." Consequently, companies like Klarna and Revolut are aspiring to become "global financial technology companies," recognizing the need for immense scale (Klarna currently has 111 million customers) to thrive in a future with "less money per customer." The strategy for investors is to aim for a "bigger piece of a smaller pie."

Personal Use of AI and Advice

The speaker personally uses AI "all the time." A recent adoption since May is "vibe coding" (AI-assisted coding) to explore codebases, despite not being an engineer by training. He finds this experience "mind blowing" and "fantastic," often spending evenings on it.

His advice is to "embrace it, learn it, educate yourself, find out how to utilize it" rather than being passively worried about technology. Understanding what one is "actually facing" helps alleviate anxiety, as fear often stems from the unknown.