Sam Altman Says People Are “Right to Be Afraid” of AI. So Why Trust AI Firms?

Artificial intelligence is advancing quickly, and the people building it are increasingly acknowledging that its risks deserve serious attention. OpenAI chief executive Sam Altman recently said that people are “right to be afraid” of AI, while also urging the public to trust AI companies to make responsible decisions.

His remarks raise an important question: If AI companies recognise that their technology could cause serious harm, should the public rely on those same companies to keep it safe?

The debate is not simply about whether AI is useful or dangerous. It is about who gets to decide how powerful AI systems are developed, tested, released, and governed.

What Did Sam Altman Say?

Speaking at Salesforce’s Dreamforce conference in San Francisco, Altman acknowledged that AI’s growing capabilities have made potential risks harder to dismiss. He said people are right to worry about the possibility of AI companies becoming too powerful and influencing the economy or public opinion.

At the same time, Altman expressed confidence that AI companies could act responsibly. He said that if safety and alignment could not remain ahead of AI capabilities, companies should be prepared to slow down or stop development.

The comments came amid wider discussions about AI safety, the pace of development, and whether governments should introduce stronger safeguards. 

Altman’s message contains two ideas that deserve separate consideration. AI risks are real, and AI companies believe they can manage those risks. The debate is about whether that confidence is enough.

Why are People Concerned about AI Companies?

AI systems are becoming more capable and are being integrated into workplaces, education, communication, and public services. As their influence grows, so does the importance of decisions made by the organisations developing them.

There are several concerns:

• Safety: Could advanced AI systems behave in unexpected ways or cause harm when used at scale?

• Concentration of power: Could a small number of companies gain too much influence over technology, information, and economic activity?

• Accountability: Who is responsible when an AI system causes harm?

• Transparency: Can users and independent experts understand how systems are tested and what limitations they have?

These concerns are not limited to hypothetical future scenarios. AI tools already influence decisions and workflows in many sectors.

Curious about what these systems can actually do? Explore how AI agents work and discover what makes them different from ordinary chatbots.

The more AI systems can influence real-world decisions and actions, the more important safety, transparency, and accountability become.

Can AI Firms Be Trusted to Regulate Themselves?

AI companies argue that they have strong reasons to develop safer systems. A serious failure could damage public trust, create legal consequences, and affect their businesses. Some industry leaders also believe safety can be addressed through engineering, testing, monitoring, and responsible release practices.

However, critics argue that commercial incentives can conflict with public interest. Companies compete to develop more capable systems, attract customers, and release products quickly. This creates a difficult question: Can companies consistently prioritise safety when they are also competing for market leadership?

The debate has led to calls for independent evaluation, clearer standards, and greater government oversight. These measures do not necessarily mean stopping AI development. They can help establish shared expectations about what responsible development should involve.

Trust is important, but it becomes stronger when it is supported by evidence, independent checks, and clear accountability.

Want to understand what could make you valuable in an AI-driven workplace? Explore AI skills and the future of work to identify where your learning journey could take you next.

What Could Responsible AI Development Look Like?

Responsible AI development requires more than a public promise to do the right thing. It needs practical measures that can be assessed.

Some important measures include:

AI safety is not only a concern for researchers or company executives. It also matters to people who use AI tools in their daily work.

For example, someone using AI to analyse information should know how to check the result. A developer building an AI application should consider how it might fail. If you want to understand how developers tackle the reliability problem, explore why RAG is becoming an essential skill for generative AI engineers. It offers a closer look at the technique and why it matters, while remembering that better context does not guarantee perfect accuracy.

What Does This Debate Mean for Learners and Professionals?

The discussion around AI safety may seem distant from someone preparing for a job or learning a new skill. In reality, it has practical implications.

As AI tools become more common, professionals may be expected to understand not only how to use them, but also when their output should be questioned.

For learners, this means developing a combination of technical understanding and critical thinking. A person working with AI-generated content should be able to verify facts. Someone using AI for data analysis should understand the underlying data. A developer should know how to test and secure an application.

AI literacy is not just about using a tool. It also means understanding its limitations and taking responsibility for how its output is applied.

If you want to move from simply using AI to understanding how to build and evaluate it, Gen AI Engineering can be a practical next step towards developing the technical skills to work with AI more confidently.

Key Takeaways

• Sam Altman’s remarks acknowledge that concerns about AI’s growing capabilities and influence are legitimate.

• Public confidence in AI companies depends on more than their assurances. Independent testing, transparency, and clear responsibility matter.

• As AI systems influence more decisions, questions about safety, oversight, and regulation will become harder to ignore.

• Understanding how to use AI is important, but knowing when to question its output is equally valuable.

• AI safety is not only a concern for developers and policymakers. It also matters to learners, professionals, businesses, and everyday users.

The Bigger Question is About Accountability

Sam Altman’s remarks highlight a tension at the centre of the AI industry. Companies developing increasingly powerful systems are asking the public to trust their judgement, while critics argue that decisions with such broad consequences should not rest with private companies alone.

The answer may not be a simple choice between trusting AI firms and rejecting their technology. The more useful question is what conditions should make that trust justified.

Can companies demonstrate that safety measures are effective? Can independent experts assess the risks? Are there clear responsibilities when systems cause harm? And can the public have a meaningful voice in decisions that affect society?

Trust in AI should not depend only on what companies say. It should also depend on what they demonstrate.