Goldman’s Solomon wants ‘clear human accountability’ on AI

• Key insight: In a call with investors, Solomon acknowledged that the buildout of AI’s infrastructure poses a variety of risks.
• Supporting data: According to a recent study, total investments in U.S. data centers are on track to exceed $10 trillion by 2032.
• Expert quote: “The choice is not between halting progress and moving forward without guardrails. We must enable innovation and also demand responsibility.” — Goldman Sachs CEO David Solomon

Processing Content

Goldman Sachs CEO David Solomon has been bullish for years about artificial intelligence, but on Tuesday he offered some of his clearest warnings yet of the dangers — and called for some man-made safety measures.

Those dangers, Solomon said during the bank’s third-quarter earnings call, include likely “bumps” in the economy, credit risks, a gap between AI supply and demand and, apart from financial concerns, the threat of AI bots breaking free of human control.

In response, he urged governments and tech companies to establish stronger safeguards.

“The choice is not between halting progress and moving forward without guardrails,” the CEO said. “We must enable innovation and also demand responsibility.”

In recent months, the size of America’s spending on the AI buildout has come into clearer focus. According to one recent study, total investments in U.S. data centers are on track to exceed $10 trillion by 2032, making up about 3.6% of U.S. gross domestic product per year.

At the same time, safety concerns about the technology itself have repeatedly made headlines. In July, a rogue OpenAI model hacked into the servers of the tech startup Hugging Face. In September, a researcher at the AI firm Anthropic abruptly resigned, warning that AI “could kill us all by the end of the decade.”

During Tuesday’s call, Solomon said Goldman recognizes these concerns, including the “unchecked advancement of frontier models.” To address them, he recommended applying some of the lessons learned in the past by the financial industry — including the need for human responsibility.

“The experience of the financial services industry has demonstrated the value of rigorous testing, independent review, continuous monitoring and clear human accountability,” he said. “These principles, applied in ways that reflect the unique characteristics of AI, can help build the trust and confidence needed to deploy the technology responsibly at scale.”

The comments represented, if not a change in tone, at least a significant shift in emphasis for Solomon, who has long expressed optimism about AI’s potential to fuel business investment. In 2024, he said the technology was creating a “very, very constructive runway of opportunity” for Goldman.

But he has also occasionally warned that the buildout of AI’s infrastructure could cause “bumps and recalibrations” in the U.S. economy. On Tuesday, while still expressing overall optimism about the technology, he stressed the potential for turbulence in much stronger terms.

“There’s a lot of demand for compute at the moment. The supply has not come into balance,” Solomon said. “There will be a point in time when this recalibrates, resets, for sure. And when that does happen, of course, there’ll be some bumps.”

Goldman’s own responsibility, the CEO said, will be to carefully manage risks as it lends to businesses involved in the buildout.

“We sit in the center of that ecosystem, but we also recognize we have an extraordinary obligation to be very disciplined about what we underwrite,” Solomon said.

The CEO’s comments came after a particularly strong quarter for Goldman. Earnings per share for the investment banking giant reached $17.01 in the third quarter of 2026, handily beating analysts’ consensus estimate of $13.09, according to S&P.

Net income for the $2.18 trillion-asset bank came out to $5.29 billion, up 29% from the third quarter of 2025. Total net revenue was $18.68 billion, up 23% from the same period last year.

Erika Najarian, an analyst at UBS, described Goldman’s results in a research note as a “strong clean beat.”.

Even amid his warnings, Solomon emphasized that the AI boom continues to offer Goldman many opportunities.

“The direction of travel, how the technology will be evolved, how it will be deployed — that’s just an unknown we have to live with,” he said. “But it’s an exciting time, because there’s going to be a lot of really interesting things to do around that.”