Sam Altman Says Society Must Accept Some AI Risks for Its Benefits- OpenAI CEO Sam Altman has argued that society will have to accept a certain level of risk from artificial intelligence if it wants to fully benefit from the technology, reigniting debate over how quickly AI companies should develop and deploy increasingly powerful systems.
In a recent interview, Altman acknowledged that the rapid expansion of AI will inevitably bring negative consequences, including hacks, scams and other forms of misuse. However, he said those risks have to be weighed against the potential benefits that AI could deliver to individuals and society.
Altman’s comments highlight a growing divide within the AI industry over how the technology should be regulated. While some researchers and companies are calling for stronger safeguards and slower development, OpenAI has generally advocated for allowing AI to advance while building protections as new risks emerge.
‘Some Bad Things’ Will Happen
Discussing OpenAI’s approach compared with more cautious AI-safety advocates, Altman said there was a significant difference in philosophy.
His argument is based on the idea that preventing every possible harmful use of AI may also restrict people from accessing its legitimate benefits. Instead, society should develop the ability to identify, contain and respond to problems as the technology becomes more widespread.
That could mean dealing with a range of challenges, from increasingly sophisticated online scams to cyberattacks and other forms of abuse involving AI-generated content.
Altman described this as part of the broader process of society developing “resilience” to AI-related risks. In his view, some level of harm is likely to occur regardless of how carefully the technology is regulated.
A Different Approach to AI Safety
The comments come amid an ongoing debate about the appropriate balance between AI innovation and safety.
Some AI-safety researchers have warned that increasingly capable systems could create risks that are difficult to control. Their concerns range from misinformation and fraud to more serious scenarios involving autonomous systems, cybersecurity and the possibility of AI capabilities developing faster than society’s ability to manage them.
Anthropic, another major AI company, has taken a more cautious public position on some aspects of AI development. Researchers associated with the company have raised concerns about the possibility of advanced AI creating severe risks if adequate safeguards are not established.
Altman’s position does not necessarily reject AI safety. Instead, it reflects a different assessment of how much risk should be tolerated while the technology continues to develop.
The Regulatory Question
One of the most important parts of Altman’s argument concerns regulation.
He has supported a lighter-touch regulatory approach, suggesting that excessive restrictions could slow innovation and prevent people from benefiting from useful AI applications.
Supporters of this approach argue that AI is already becoming an important tool in areas such as education, software development, scientific research, healthcare and business. They contend that regulations need to evolve alongside the technology rather than attempting to anticipate and prohibit every possible problem in advance.
Critics, however, argue that waiting for society to adapt could mean real people bear the cost of failures before adequate protections are established.
That disagreement is likely to become increasingly significant as AI systems become more capable and are integrated into everyday services.
OpenAI Faces Its Own Safety Debate
Altman’s comments also come at a sensitive time for OpenAI, with questions being raised about the company’s internal safety culture.
The discussion follows the departure of an OpenAI safety expert who criticized the company’s approach, saying its “culture is broken.” The resignation has added to broader questions about whether AI companies can simultaneously pursue rapid commercial growth and maintain strong internal safety standards.
For OpenAI, the challenge is particularly significant because its technology is being used by millions of people and its decisions can influence how the wider AI industry approaches deployment.
What Happens Next?
The central question is no longer simply whether AI will create risks. Most experts and technology companies already acknowledge that it will.
The more difficult question is how society should respond.
Should governments attempt to prevent potentially harmful AI applications before they become widespread? Or should companies be allowed to continue developing the technology while governments and users gradually adapt to its risks?
Altman’s comments suggest that OpenAI favors the second approach, with safeguards and regulation evolving alongside technological progress.
The debate is unlikely to disappear soon. As AI becomes more powerful and more deeply embedded in everyday life, society will have to decide which risks are unacceptable, which can be managed, and how much uncertainty it is willing to tolerate in exchange for technological progress.
For Altman, the potential benefits of AI make accepting some level of risk a necessary part of moving forward.
