OpenAI CEO Sam Altman Warns of AI Risks, Announces New Role to Tackle Security Threats

OpenAI CEO Sam Altman has signaled growing concerns over the potential dangers of artificial intelligence (AI). In a post on social media platform X, he announced the creation of a new position, Head of Preparedness, aimed at addressing security risks associated with AI agents and prompt injection. The role carries a salary of $555,000 along with equity incentives.

Addressing Emerging AI Threats

Altman highlighted that while AI models are advancing rapidly and capable of performing remarkable tasks, they also present real-world challenges. “It is crucial that the power of these models is used responsibly. We must ensure they remain safe and are not misused,” he wrote. He warned that the security risks posed by AI agents could become more significant in 2026, particularly as models are increasingly capable of identifying serious vulnerabilities in computer systems.

AI Agents and Prompt Injection

The CEO’s post underscores OpenAI’s vigilance regarding AI agents that can autonomously execute tasks and the risks of prompt injection, a scenario where hackers manipulate AI models by providing malicious instructions. Such vulnerabilities pose serious threats to cybersecurity.

Why This Step Was Taken

OpenAI is seeking professionals to manage these emerging risks, reflecting the dual nature of AI: rapid technological advancement alongside escalating security challenges. The move follows observations in 2025 of AI models impacting mental health and detecting critical system weaknesses. Recent reports also indicate that AI systems are being used in cyberattacks. Last month, Anthropic revealed hackers exploited their cloud-based tools to target nearly 30 global organizations, including tech companies, financial institutions, and government agencies.

Ongoing Legal Concerns

Altman also raised concerns over AI’s potential effects on mental health. OpenAI is already facing multiple high-profile lawsuits, including cases alleging ChatGPT’s role in teen suicides. Reports of AI chatbots misleading users have also emerged, prompting the company to proactively prepare for these challenges.

With this new preparedness initiative, OpenAI aims to stay ahead of security threats, ensuring AI technologies remain safe, ethical, and beneficial while mitigating risks to users and systems worldwide.