By Isnaini Azizah 
Matthew Mish, the head of credit strategy at UBS, warns that tens of billions of dollars in corporate loans are likely to default over the next year. The primary victims of this shift are expected to be software and data services firms, particularly those owned by private equity firms and burdened with high levels of debt. These companies, which once enjoyed stable cash flows and predictable growth, now find themselves squeezed between the rising costs of servicing their debt and the existential threat posed by rapid AI advancements. Mish’s research note, released Wednesday, highlights a growing concern that the market has underestimated the velocity of AI-driven disruption, leading to a "rapid, aggressive disruption scenario" that is no longer a distant theoretical possibility but a looming reality for 2026 and beyond.
The urgency of this recalibration stems from the blistering pace of technological development. Mish noted that he and his team at UBS were forced to accelerate their forecasts following the release of the latest large language models from industry leaders like Anthropic and OpenAI. These new iterations have demonstrated capabilities that threaten to render traditional software architectures and data processing workflows obsolete much faster than previously anticipated. "The market has been slow to react because they didn’t really think it was going to happen this fast," Mish told CNBC in a recent interview. He emphasized that investors and credit analysts must stop viewing AI disruption as a problem for 2027 or 2028; rather, it is a structural shift currently manifesting in the credit markets.
The shift in investor sentiment has been palpable. Earlier this year, AI was viewed as a "rising tide" that would lift all technology stocks. However, as the market matures, a "winner-take-all" dynamic has emerged. While foundational model creators like OpenAI and Anthropic are seeing their valuations skyrocket—with Anthropic recently closing a $30 billion funding round—incumbent software firms are facing intense scrutiny. This rolling series of selloffs has not been confined to the tech sector alone. In recent weeks, the "AI contagion" has spread to finance, real estate, and even trucking, as investors realize that automation and intelligent agents could drastically reduce the need for human labor and physical office space, thereby undermining the collateral value of many corporate loans.
In his baseline scenario, Mish and the UBS credit strategy team project a significant spike in defaults among borrowers of leveraged loans and private credit. Specifically, they estimate that between $75 billion and $120 billion in fresh defaults could hit these markets by the end of this year. This calculation is rooted in the sheer size of the "shadow banking" and leveraged finance sectors. Mish estimates the leveraged loan market to be approximately $1.5 trillion, while the private credit market—often less transparent and more aggressive in its lending practices—has ballooned to $2 trillion. A projected increase of 2.5% in defaults for leveraged loans and 4% for private credit by late 2026 would result in a massive destruction of capital, potentially destabilizing the financial institutions that hold this debt.
However, the "baseline" scenario may be optimistic. Mish also highlighted a "tail risk" scenario involving a more sudden and painful AI transition. In this environment, defaults could jump by twice the baseline estimates, effectively cutting off funding for a wide swath of middle-market companies. This would lead to what Wall Street jargon terms a "credit crunch." If the loan markets reprice aggressively, the cost of capital would soar, creating a feedback loop where even fundamentally sound companies struggle to refinance their obligations. "The knock-on effect will be that you will have a credit crunch in loan markets," Mish warned. "You will have a broad repricing of leveraged credit, and you will have a shock to the system coming from credit."
The vulnerability of these markets is exacerbated by the way corporate debt has been structured over the last decade. Many software and data services firms were acquired by private equity firms during a period of ultra-low interest rates. These deals were often financed with high levels of floating-rate debt, meaning that as central banks raised rates to combat inflation, the cost of servicing that debt skyrocketed. Now, these same companies are facing a decline in revenue or a compression of margins as AI-native startups offer faster, cheaper, and more efficient alternatives to their legacy products. Without the cash flow to both pay down debt and invest in their own AI transformations, these firms are caught in a "debt trap" that leaves them vulnerable to insolvency.
To better understand the landscape of AI risk, Mish categorizes companies into three distinct buckets. The first category comprises the "Foundational Kings"—the creators of the large language models (LLMs) like OpenAI and Anthropic. These firms are currently the primary beneficiaries of the AI boom, attracting massive capital infusions and setting the pace for the rest of the industry. While they are currently startups, they are quickly evolving into the next generation of tech giants, potentially displacing the very companies that currently dominate the indices.
The second category includes "Investment-Grade Adaptors," such as Salesforce and Adobe. These are established software giants with robust balance sheets, massive R&D budgets, and deeply entrenched customer relationships. Unlike their smaller, more indebted peers, these companies have the financial firepower to integrate AI into their existing ecosystems, thereby fending off challengers and maintaining their market positions. For these firms, AI is an opportunity to upsell existing clients and improve internal efficiencies, provided they can innovate fast enough to keep pace with the foundational model creators.
The third and most at-risk category is the cohort of private equity-owned software and data services companies that carry high debt-to-equity ratios. These firms often lack the liquidity to pivot their business models. "The winners of this entire transformation—if it really becomes, as we’re increasingly believing, a rapid and very disruptive or severe change—the winners are least likely to come from that third bucket," Mish said. For many of these companies, the "creative destruction" promised by AI may lean more toward the "destruction" side of the equation, as their legacy value propositions are eroded by automation.
The broader implications of this credit risk extend into other sectors of the economy. For instance, office real estate stocks have already begun to tumble as investors anticipate that AI-driven productivity gains will lead to smaller corporate headcounts and a reduced need for physical workspace. Similarly, the trucking and logistics sectors are bracing for the arrival of autonomous systems that could disrupt the traditional labor-intensive model. If the credit markets for these sectors freeze up, the resulting economic slowdown could be more severe than a typical cyclical downturn.
Furthermore, the "private credit" boom of the last few years is facing its first true stress test. Unlike traditional bank loans, private credit is often held by non-bank financial institutions, insurance companies, and pension funds. Because these loans are not publicly traded, their true value can be obscured until a default occurs. If Mish’s "tail risk" scenario manifests, the lack of transparency in private credit could lead to a sudden and unexpected wave of write-downs, catching institutional investors off guard and potentially leading to systemic instability.
As the market moves toward this "rapid disruption" phase, the timing of AI adoption by large corporations remains the critical variable. While the technology is advancing at a breakneck pace, the integration of AI into complex corporate workflows often takes longer than the "hype cycle" suggests. However, Mish argues that the window for adaptation is closing. The "recalibration" he speaks of is not just about changing stock price targets; it is about a fundamental reassessment of creditworthiness in an era where software can be generated by machines and data can be processed without human intervention.
Ultimately, the warning from UBS serves as a harbinger of a new era in financial risk management. For years, the primary concern for credit investors was interest rate risk and the traditional business cycle. Now, "disruption risk" has moved to the forefront. As AI continues to evolve from a buzzword into a structural economic force, the divide between the "AI-enabled" and the "AI-disrupted" will likely be written in the ledgers of the credit markets, with tens of billions of dollars hanging in the balance. The "credit crunch" Mish envisions may be the ultimate price the market pays for the unprecedented speed of the artificial intelligence revolution.