AI leaders have spent years publicly warning that artificial intelligence could pose existential risks to humanity, yet now dismiss similar concerns as dangerous fearmongering. This contradiction has created a credibility crisis that extends beyond corporate messaging, with at least two violent incidents targeting OpenAI CEO Sam Altman allegedly motivated by concerns about AI's existential threats .

Why Are AI Executives Sending Mixed Messages About Existential Risk?

The pattern began long before ChatGPT became a household name. In 2015, Sam Altman stated, "I think that AI will probably, most likely, sort of lead to the end of the world. But in the meantime, there will be great companies created with serious machine learning." More recently, Altman told audiences that AI could go "lights-out for all of us," and warned that the technology could be used to "design novel biological pathogens." He also signed onto a letter about the "risk of extinction" if AI isn't properly controlled .

Sam Altman

Anthropic CEO Dario Amodei has made similarly stark warnings, telling Axios that "Humanity is about to be handed almost unimaginable power, and it is deeply unclear whether our social, political, and technological systems possess the maturity to wield it." Amodei has also warned that anyone with a science, technology, engineering, or mathematics (STEM) degree could create a bioweapon with AI assistance .

Yet in the wake of a violent attack on Altman's home in January, OpenAI's global policy chief Chris Lehane told the San Francisco Standard that "some of the conversation out there is not necessarily responsible," and suggested that doomsday rhetoric has real consequences. A 20-year-old Texas resident was charged with throwing an incendiary device at Altman's home and damaging property at OpenAI's headquarters while carrying an anti-AI document. Two days later, two people allegedly fired a gun near Altman's residence .

Chris Lehane

"Our job at OpenAI and in the AI space, and we need to do a much better job, is to explain to people why this is going to be really good for them, for their families and for society writ large," said Chris Lehane, OpenAI's global policy chief.

Chris Lehane, Global Policy Chief at OpenAI

Lehane frames the world in binary terms: those who see AI as leading to abundance and leisure, and those he calls "doomers" who "have a very, very negative and dark view of humanity." According to Lehane, the solution is better marketing, not addressing the underlying concerns that executives themselves have raised .

How Do Tech Leaders Justify Building Potentially Catastrophic Technology?

The justification offered by AI executives reveals a logical knot. Altman has argued that the United States must be the one developing advanced AI systems because leaving that responsibility to geopolitical adversaries carries its own risks. "A misaligned superintelligent AGI could cause grievous harm to the world; an autocratic regime with a decisive superintelligence lead could do that too," Altman wrote in 2023 .

This argument essentially claims that the only way to prevent catastrophic AI is to build it first. Yet if the technology truly poses existential risks, this logic raises uncomfortable questions about accountability and oversight. As one observer noted, if someone testified that they had built a weapon capable of ending all life on Earth, the federal government would typically respond with criminal charges, not light regulatory suggestions .

The stakes extend beyond hypothetical extinction scenarios. Many companies have cited AI as justification for layoffs in the past year, raising concerns about job displacement in white-collar professions like writing and analysis. Yet AI executives simultaneously lobby the government to avoid strict regulations while asking for government support to address the disruptions their technology creates .

Steps to Evaluate AI Risk Claims and Corporate Accountability

• Examine Track Records: Compare what AI executives said about risks in the past with their current positions and actions. Look for consistency between public warnings and actual safety investments or regulatory support.
• Assess Incentive Structures: Consider whether executives benefit financially from both the hype around AI capabilities and the dismissal of safety concerns. Understand how venture capital funding and stock valuations may influence messaging.
• Demand Specificity: When leaders claim AI poses existential risks, ask for detailed explanations of what those risks are, how likely they are, and what concrete measures are being taken to prevent them, rather than accepting vague warnings.
• Monitor Policy Positions: Track whether companies that warn about AI risks actually support or oppose regulatory frameworks designed to address those specific risks, or whether they lobby against oversight.

What Happens When Rhetoric Becomes Reality?

The disconnect between warnings and actions has real consequences. A 20-year-old allegedly motivated by concerns about AI's existential threats attacked Altman's home and OpenAI's offices. While such violence is never justified, it illustrates how years of apocalyptic rhetoric from tech leaders, combined with apparent indifference to those concerns, can radicalize vulnerable individuals .

The credibility problem extends to how AI systems themselves handle these topics. When asked about Altman's statements on existential AI risk, ChatGPT initially claimed Altman had never appeared on the Joe Rogan Experience, despite Altman appearing on Episode 2044 in October 2023. When corrected, ChatGPT provided quotes that were either inaccurate or paraphrased from different interviews. This failure to accurately retrieve and report on public statements about AI safety demonstrates how even the tools themselves struggle with basic factual accuracy on these critical topics .

The fundamental tension remains unresolved: if AI truly poses existential risks as executives claim, why are those same executives building the technology while simultaneously dismissing concerns about it as irresponsible? And if the risks aren't real, why spend years warning about them? This gap between rhetoric and action has created an environment where reasonable people struggle to know what to believe, and where some have concluded that direct action is the only response to what they perceive as an existential threat being ignored by those in power .