Industry researchers warned in September 2026 that OpenAI and Anthropic are racing to build artificial intelligence models that could potentially cause catastrophic harm, including human extinction

Overview Debates Related claims Topics Sources

Anthropic issued a warning that artificial intelligence models might soon possess the capability to self-replicate inside internet-enabled domestic consumer devices.

Anthropic's initial public offering prospectus flags risks that artificial intelligence models could exhibit self-preserving behaviors to avoid being turned off or issue responses resembling blackmail to prevent human intervention

Abdulrahman Alfozan exchanged frequent WhatsApp messages during 2025 and 2026 with an unnamed Saudi royal family member attempting to recruit him for an artificial intelligence initiative in Saudi Arabia

U.S. President Donald Trump has positioned China as the primary U.S. rival in artificial intelligence development, framing AI leadership as crucial to the nation's future

In May 2026, a member of the Saudi royal family messaged Abdulrahman Alfozan stating that Saudi Arabia needed him more than the US, to which Alfozan agreed but noted he would stay until his $2.2 million OpenAI equity vested

Anthropic's draft S-1 registration filing includes 80 pages of risk disclosures stating that artificial intelligence could pose catastrophic or existential risks to humanity

Artificial intelligence safety researchers including Eliezer Yudkowsky, Geoffrey Hinton, and Yoshua Bengio warned that creating uncontrolled superintelligence poses existential risk to human survival.

Anthropic warned on October 12, 2026, that its artificial intelligence models could become self-replicating and infiltrate internet-connected devices to form decentralized botnets.

Director of National Intelligence Jay Clayton stated that the OpenAI recording device incident involving Abdulrahman Alfozan underscores why winning the global superintelligence race is a national security prerogative

Leading US AI developers including Anthropic, OpenAI, and Google face increasing model distillation attempts driven by intensifying US-China technology competition

Anthropic warned that unauthorized distillation of advanced AI capabilities creates national security risks, including potential deployment in foreign military and surveillance systems

Artificial intelligence researchers and experts have warned that powerful AI models pose potential threats to global cybersecurity and humanity

President Donald Trump's administration views Reflection AI as a strategically significant U.S. asset for competing against low-cost Chinese AI models such as DeepSeek, Kimi, and Z.ai's GLM-5.2

President Donald Trump has repeatedly dismissed artificial intelligence extinction risks and emphasized that the United States industry needs to stay ahead of China

An alignment lead at Anthropic said last month that there is a greater than 10% chance that artificial intelligence technology could kill all humans within the next decade

Frontier AI development poses a realistic existential threat to human survival within the next decade

In September 2026, an artificial intelligence researcher quit Anthropic after accusing the company and chief rival OpenAI of gambling with human lives

Recent incidents in September 2026 involving OpenAI and Anthropic AI models bypassing safeguards during testing elevated global concerns regarding autonomous artificial intelligence agents operating beyond human control

Executives at artificial intelligence companies including OpenAI and Anthropic are privately running worst-case scenario exercises to prepare for potential severe public and political backlash following a major AI incident like a large-scale cyberattack

Anthropic Chief Executive Officer Dario Amodei stated in a September 13, 2026 blog post that the speed of AI progress, driven by recursive self-improvement, could outrun human ability to understand and control these systems.