In the ongoing negotiation between human ambition and human welfare, OpenAI's Sam Altman has staked a position as old as industry itself: that transformative progress demands a tolerance for harm. Speaking publicly in October 2026, Altman argued that AI's promised benefits justify accepting certain negative consequences along the way — a calculus that other technology leaders have been quick to challenge. The exchange reveals something deeper than a policy dispute: it is a question about who holds the authority to decide which risks are worth taking, and on whose behalf.
Altman: AI's benefits justify accepting some risks and negative outcomes
The world should accept some bad things happening for AI's benefits
Mark
So Altman is saying we should just accept harm from AI as the cost of progress. That's a pretty bold thing to say out loud.
Mimi
It is, but it's also the logic that's been driving the industry for a while now. He's just naming it explicitly. The idea is that you can't build transformative technology without some friction, some unintended consequences.
Luke
But here's what I want to know: when he says "some bad things," what does that actually mean? Is it job losses? Discrimination? Security breaches? The statement is vague enough that it could mean almost anything.
Mimi
That's fair. The statement doesn't specify. But I think the underlying claim is that the benefits—medical advances, productivity gains, solutions we haven't even imagined yet—are worth the risks we can't fully predict.
Mark
And the Zscaler CEO disagreed with that framing entirely.
Luke
Right, but we should be careful here. One CEO disagreeing with another CEO is not the same as a broad consensus that Altman is wrong. We don't know how many other tech leaders agree with him versus how many don't. The reporting shows us a disagreement exists, but not its scale.
Mimi
That's true. But the disagreement does point to something real: there's no consensus in the industry about how much risk is acceptable. Some people think you iterate and manage problems as they come. Others think you need stronger safeguards before deployment.
Mark
Who bears the cost if things go wrong?
Mimi
That's the hard part. Altman speaks from a position of power and wealth. The people who might lose jobs to AI, or be harmed by algorithmic bias, or have their privacy violated—they didn't choose to accept that risk. It was chosen for them.
Luke
And we should note: we don't actually know yet what the full range of harms might be. AI is still relatively new in its current form. So when Altman says we should accept bad outcomes, we're accepting outcomes we can't fully predict or measure.
O Pulso
- Altman's declaration that society must accept 'some bad things happening' as the cost of AI progress has ignited a sharp and widening debate across the technology industry.
- Critics, including Zscaler's CEO, have publicly rejected the framing — arguing that tolerating AI-related harms is not an inevitable condition of progress but a choice being made by a powerful few on behalf of everyone else.
- At the heart of the tension is a profound asymmetry: those deciding to absorb the risk are not the same people who will bear its consequences — job losses, algorithmic bias, and privacy violations fall on individuals, not executives.
- Regulation and public understanding continue to lag far behind deployment, leaving the definition of 'acceptable harm' undefined and contested, with no shared framework yet in place.
- The debate is landing in an unresolved space — neither side has won the argument, and the industry continues to move forward while the terms of the trade-off remain unanswered.
In the ongoing negotiation between human ambition and human welfare, OpenAI's Sam Altman has staked a position as old as industry itself: that transformative progress demands a tolerance for harm. Speaking publicly in October 2026, Altman argued that AI's promised benefits justify accepting certain negative consequences along the way — a calculus that other technology leaders have been quick to challenge. The exchange reveals something deeper than a policy dispute: it is a question about who holds the authority to decide which risks are worth taking, and on whose behalf.
Sam Altman, chief executive of OpenAI, has publicly argued that advancing artificial intelligence requires the world to accept certain negative consequences. Speaking to the outlet Decoded, he framed the position plainly: some bad things will happen, but the broader benefits of AI — medical breakthroughs, productivity gains, solutions to problems not yet solved — justify tolerating those harms. The logic echoes a familiar defense of transformative technology: waiting for perfect safety means waiting forever.
But the statement crystallized a discomfort many observers have long carried about the current pace of AI development. The harms Altman describes as abstract trade-offs — job displacement, algorithmic discrimination, misinformation — are concrete realities for the individuals who experience them. The asymmetry is difficult to ignore: those with the power to decide the pace of deployment are not the same people who absorb its costs.
Pushback arrived quickly. Zscaler's CEO publicly rejected Altman's framing, insisting that society should not simply accept AI-related harms as the price of progress. The disagreement reflects a genuine fault line in the industry — between those who see safety concerns as manageable through iteration and those who believe certain risks demand stronger guardrails before wider deployment.
What remains unresolved is the substance behind Altman's words. How much displacement is acceptable? What degree of bias is tolerable? At what point do accumulated small harms become a large one? These questions have no clear answers, and Altman's statement does not attempt to provide them. The confidence he projects — that the benefits will justify the costs — is not yet shared universally, and the debate over who gets to make that determination on society's behalf is far from settled.
Sam Altman, the chief executive of OpenAI, has argued publicly that the world must accept certain negative consequences as the price of advancing artificial intelligence. Speaking to the outlet Decoded, Altman framed the trade-off plainly: some bad things will happen, he said, but the benefits that AI promises justify tolerating those harms. The statement landed in a moment of intense debate about how quickly AI should be deployed into society and who bears the cost when things go wrong.
Altman's position reflects a calculation that has become central to the AI industry's push for rapid development. The logic is straightforward enough—transformative technologies always carry risks, the argument goes, and waiting for perfect safety is waiting forever. Move forward, learn as you go, and manage problems as they emerge. From this view, the alternative to accepting some damage is forgoing the gains entirely: medical breakthroughs, productivity improvements, solutions to problems we haven't yet solved. The framing puts the burden of proof on skeptics to explain why AI should be treated differently from every other powerful tool humanity has ever built.
But the statement also crystallized something that has troubled many observers about the current trajectory of AI development: the asymmetry between who decides to take the risk and who bears the consequences. Altman speaks from a position of enormous influence and wealth. The "bad things" he references are abstract at the level of corporate strategy but concrete at the level of individual lives—job displacement, algorithmic discrimination, misinformation, privacy violations, or worse harms not yet fully understood.
The pushback came swiftly. Zscaler's CEO publicly rejected Altman's framing, arguing that society should not simply accept AI-related harms as inevitable. This disagreement is not merely philosophical. It reflects a genuine fault line in how different leaders in technology believe the industry should operate. Some, like Altman, see safety concerns as constraints that slow progress and should be managed through iteration. Others believe that certain risks are too large or too poorly understood to proceed without stronger guardrails in place first.
The debate also touches on questions of power and accountability. OpenAI and a handful of other companies control the most advanced AI systems in existence. They make decisions about how these systems are trained, deployed, and monitored. When Altman says the world should accept bad outcomes, he is, in effect, saying that OpenAI and companies like it should be permitted to proceed with their work even when the full consequences remain unknown. The world, in this formulation, does not get to vote—it gets to absorb the results.
What remains unresolved is what "some bad things" actually means in practice. How many job losses constitute an acceptable cost? What level of algorithmic bias is tolerable? At what point does the accumulation of small harms become a large one? These are not questions Altman's statement addresses. They are also not questions that have clear answers, which is partly why the debate persists. The tech industry has moved faster than regulation, faster than public understanding, and faster than the development of shared norms about what kinds of risks are worth taking on behalf of society as a whole. Altman's comment reflects confidence that the benefits will justify the costs. The disagreement from other leaders suggests that confidence is not yet universal.
Citações Notáveis
The world should accept some bad things happening as a necessary cost of advancing AI technology— Sam Altman, OpenAI CEO, speaking to Decoded
Society should not simply accept AI-related harms as inevitable— Zscaler CEO, in public disagreement with Altman
Quer a matéria completa?
Leia o original em Google News ↗
Technology