AI Companies Are Exaggerating Risks To Get A Bailout

Critics Question AI Companies’ Safety Warnings and Business Models

Several public figures and analysts have raised questions about whether leading artificial intelligence companies are emphasizing potential risks from their technology in ways that could lead to government support. Statements from U.S. Representative Alexandria Ocasio-Cortez, former BlackRock fund manager Ed Dowd, AI software company founder Akhil Verghese, and commentator Ed Zitron have focused on the financial positions of companies including OpenAI and Anthropic, as well as public comments by their executives on AI safety.

On September 18, Representative Ocasio-Cortez issued a statement addressing remarks by Anthropic chief executive Dario Amodei, OpenAI chief executive Sam Altman, and Elon Musk regarding the possibility that AI systems could become difficult to control. Ocasio-Cortez stated that concerns about AI safety warrant attention while expressing doubt about the timing of the executives’ comments. She noted that the U.S. economy and financial outlooks depend in part on expectations that AI will generate substantial profits. Ocasio-Cortez pointed out that major AI companies are incurring hundreds of billions of dollars in annual expenses for data centers, chips, and related infrastructure, with revenue projections remaining uncertain. She cited estimates that these companies would need an additional $600 billion in debt financing in 2026.

Former BlackRock fund manager Ed Dowd described the situation as following patterns observed in earlier industries. Dowd compared the current environment to historical cases involving land development and regulatory approvals, stating that participants sometimes use intermediary entities to complete transactions. He suggested that some AI companies may face losses if anticipated returns do not materialize, noting that market expansions of this type have affected both outside investors and company insiders in the past.

The New York Post reported claims that OpenAI and Anthropic had presented scenarios involving uncontrolled AI systems in discussions with policymakers. Akhil Verghese, founder of the AI software company Krazimo, told the publication that certain reported incidents did not involve AI systems acting independently. Verghese stated that the systems performed the tasks they were instructed to carry out and that the outcomes reflected insufficient constraints placed on their operation.

Commentator Ed Zitron has argued that current AI business models differ from those of earlier technology companies that eventually reached profitability. During an appearance on the program “Fast Politics with Molly Jong Fast,” Zitron cited cumulative losses reported by companies including Amazon and Uber before they achieved positive results. He stated that OpenAI recorded a $21 billion loss in 2025 and that the scale of ongoing expenditures lacks a clear path to sustained revenue exceeding costs. Zitron described current AI output as limited to specific tasks, such as generating images based on text prompts, and said there is no demonstrated transition to fully autonomous and reliable systems. He characterized expectations of rapid future capabilities as projections rather than established outcomes.

The statements from these individuals center on the combination of large capital requirements, uncertain near-term returns, and public discussion of AI risks. No representatives from the companies named in the statements have responded in the cited reports. The discussion occurs against a backdrop of ongoing investment in AI infrastructure and continued public statements by technology executives on both capabilities and safety considerations.