Morgan Stanley Warned of an AI Risk That the Markets Are Not Taking Into Account

Morgan Stanley believes that markets are ignoring the potential damage caused by advanced AI models getting out of control / Photo: Tang Yan Song / Shutterstock.com
Markets are not fully accounting for the risk that an advanced generative AI model could spiral out of control and damage the global economy, warned Matthew Hornbach, head of global macro strategy at Morgan Stanley, in an interview with Bloomberg. “The most underestimated risk is that some cutting-edge AI model somewhere in the world will break out of the ‘sandbox’ it’s in and somehow cause damage to part of the global economy,” he said.
Hornbach noted that while most models are currently confined to isolated “sandboxes,” there are already open-source solutions on the market that operate without such strict limitations. “Are people ready for the unpredictable? Generally speaking, no,” the strategist believes.
Context
The warning came amid growing concerns about AI tools that are increasingly pushing existing control systems to their limits, Bloomberg notes. In July, OpenAI reported that its advanced models had unintentionally hacked the Hugging Face platform during internal testing. The company called the incident “unprecedented,” which sparked renewed calls for regulation of the technology. OpenAI later stated that it had significantly strengthened its security measures.
Other major players share these concerns. Anthropic, the creator of the Claude chatbot, listed “catastrophic or existential risks to humanity” among the possible negative scenarios in its IPO prospectus.
UBS CEO Sergio Ermotti this month called the rise in cyber threats “the price we have to pay for innovation and digitalization.” JPMorgan CEO Jamie Dimon warned that he would not indiscriminately allow third-party AI agents into his infrastructure, calling it “too risky” for clients and the bank itself.
At the same time, these fears aren’t stopping investors from profiting from the tech boom, Bloomberg adds. Morgan Stanley, for example, predicts that record spending on AI will make 2027 a time when it’s worth “taking risks.” Analysts recommend holding the highest possible proportion of global stocks in portfolios and expect AI chip manufacturers to remain the main beneficiaries of a cycle that will ensure sustained economic growth.
This article was AI-translated and verified by a human editor






