Shares of Uber, DoorDash, Mastercard, and American Express all declined sharply this week following the circulation of a detailed analysis outlining a potential economic disruption driven by advances in artificial intelligence. The scenario, initially published on the Substack platform by researcher and analyst, Citrini, posits a future where increasingly capable AI agents fundamentally alter economic structures, leading to widespread job displacement and financial instability.
The core of Citrini’s analysis centers on the recent “jump in capability” observed in AI agents, specifically citing Anthropic’s Claude Code and OpenAI’s Codex. These tools have demonstrated an ability to perform complex tasks, such as coding and data analysis, with a level of proficiency that threatens to automate numerous white-collar jobs. According to the analysis, this automation will not follow the traditional pattern of technological advancement, where new jobs are created to offset those lost. Instead, AI’s capacity for self-improvement means displaced workers will struggle to find alternative employment within the AI sector itself.
The report suggests that the proliferation of these AI agents will significantly impact software-as-a-service (SaaS) companies like Monday.com, Zapier, and Asana. Businesses may opt to develop in-house solutions using AI, rather than subscribing to these services, driving down prices and creating a “race to the bottom” for companies like Oracle that rely on long-term contracts. This trend, Citrini argues, extends beyond the software industry. The analysis predicts that consumers will increasingly utilize personal AI agents to handle transactions, effectively eliminating the need for intermediaries like travel and estate agencies.
The disruption extends to consumer-facing services. Citrini envisions a scenario where individuals code their own food delivery applications, fragmenting the market and eroding the margins of companies like DoorDash. Similarly, ride-sharing services like Uber could see business evaporate as AI agents facilitate direct transactions between riders, and drivers. Even the payments industry is not immune, with AI agents potentially favoring cryptocurrency transactions due to lower costs, thereby undermining traditional payment providers like Visa and Mastercard.
The analysis further predicts a ripple effect through the financial system. Private credit firms, which have been actively involved in restructuring software businesses, face potential defaults as the revenue assumptions underpinning their loans grow unsustainable. Citrini specifically cites the 2022 acquisition of Zendesk by Hellman &. Friedman and Permira as an example, noting that the deal was structured around expectations of stable revenue growth – an assumption now called into question. This could trigger “the largest private credit software default in history,” potentially impacting life insurance policies and the savings of American households.
Adding to the financial strain, the analysis foresees a mortgage crisis as white-collar workers lose their jobs and struggle to meet their mortgage obligations. This combination of factors, Citrini argues, will create a negative feedback loop: layoffs lead to decreased consumer spending, prompting companies to invest further in AI and exacerbate job losses. This cycle, the analysis contends, lacks a “natural brake.”
The scenario culminates in a projected market crash in late 2027, driven by the mortgage market, wiping out 57% of the S&P 500. Citrini suggests that traditional financial policy tools will be ineffective in addressing this crisis, as it stems not from financial conditions but from the fundamental shift in the value of human labor due to AI. The resulting economic landscape will be characterized by “ghost GDP” – economic output that appears in national accounts but does not translate into real-world income or consumption.
The analysis concludes with a warning that this is the first time in history that the most productive asset in the economy – AI – has produced fewer, not more, jobs. Citrini argues that existing economic frameworks are inadequate to address this new reality, and that the development of new frameworks is urgently needed. As of Tuesday, Anthropic announced new multiyear partnerships with four major consulting firms to deploy its Frontier AI alongside OpenAI engineers, further accelerating the integration of AI into the workplace.
Meanwhile, Anthropic revoked OpenAI’s API access to its Claude models on August 2, 2025, citing violations of its terms of service. According to Anthropic, OpenAI was using Claude to benchmark its own AI models and evaluate safety features ahead of the release of GPT-5. OpenAI disputed the claim, stating that It’s standard practice to evaluate other AI systems and that its API remains available to Anthropic.
