Sen. Adam Schiff (D-Calif.) warned that increasingly advanced artificial intelligence systems pose a serious national security risk, arguing that voluntary commitments from technology companies are not enough to address potential threats.
Schiff Calls for AI Guardrails
On Wednesday, Schiff wrote on X, "When AI undermines our national security, a pinky swear won’t cut it."
He added, "We need commonsense guardrails before it’s too late."
In a video shared with the post, Schiff criticized technology leaders who signed onto a voluntary AI pledge while separately calling for regulation.
"It was interesting to see all those tech leaders there, signing onto this voluntary pledge," he said, describing the situation as very disingenuous and "very jarring."
Schiff said increasingly advanced AI models could create growing national security concerns as they improve themselves through "recursive AI."
"We have less and less of a sense of what they’re doing, how they’re deciding, what they’re deciding, what they’re trying to access lawfully or unlawfully," he said.
Schiff Raises Pentagon AI Concerns
Schiff also warned about the "unbridled implementation of AI at the Pentagon," saying it could result in "calamities."
He said there should be an investigation into what role AI may have played in target selection during the bombing of an Iranian girls’ school.
Schiff also pointed to Anthropic’s opposition to using AI for domestic mass surveillance and fully autonomous weapon systems, calling those issues a national security concern.
AI Safety and National Security Concerns
Last week, OpenAI CEO Sam Altman warned against treating AI like a deity and said replacing human judgment with AI models posed a "real safety issue."
His comments followed reports of Anthropic discussing AI consciousness and morality with religious scholars.
Last month, Microsoft Corp. (NASDAQ:MSFT) co-founder Bill Gates warned that increasingly capable AI could enable sophisticated cyberattacks and dangerous biological research, including through the "dual use" of molecular research.
Rep. Ro Khanna (D-Calif.) said he plans to introduce the Human Control Over AI Act, a proposed measure that would impose strict liability standards and restrict AI models from recursively improving themselves or autonomously modifying their core objectives, containment measures or shutdown controls until federal safeguards are established. The proposal would also create a federal agency to oversee frontier AI systems.
In July, Sen. Rick Scott (R-Fla.) warned that AI was creating new national security threats and proposed the Generative AI Terrorism Risk Assessment Act, which would require regular federal reports on how terrorist organizations were using generative AI.
Disclaimer: This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors.
Photo courtesy: Shutterstock
Login to comment