Ro Khanna Proposes Federal AI Agency to Regulate AI Safety Amid Growing Existential Risks

Ro Khanna Pushes for FDA-Style AI Agency

Rep. Ro Khanna has called for the creation of a federal AI agency modeled after the Food and Drug Administration to regulate artificial intelligence development1

1

. Speaking on CBS's Face the Nation, the California Democrat argued that AI regulation should mirror existing frameworks for nuclear energy, electricity, and aviation, with independent technical experts ensuring AI safety2

2

. Khanna dismissed claims from tech CEOs that AI governance is too complex for Washington, stating that the government has successfully regulated intricate systems before and can appoint qualified experts from universities to oversee AI development1

1

.

Comparing AI Development to Nuclear Threats

Khanna drew a stark comparison between unchecked AI development and nuclear warfare, warning that "winning an AI war is like winning a nuclear war" and would be "catastrophic for civilization"1

1

. He emphasized two primary dangers: loss of control over AI systems and deliberate misuse of the technology1

1

. The congressman has already sent letters to AI companies including DeepSeek and Alibaba, urging them to slow technological advancement until proper guardrails are established1

1

. His proposal includes mandatory safety certification, insurance requirements, and liability rules for AI systems2

2

.

International AI Agreements and China

Khanna is advocating for international AI agreements with China, specifically targeting recursive self-improving AI and safeguards against misuse for biological or nuclear weapons2

2

. He stressed the need for an international agreement with Chinese President Xi Jinping to establish basic standards including international monitoring and data center checks1

1

. The congressman acknowledged that trust verification would be essential, comparing the approach to arms control agreements and invoking Ronald Reagan's "trust but verify" principle1

1

. This push comes ahead of Xi's expected meeting with President Donald Trump at Joint Base Andrews for a state visit1

1

.

Trump Administration's Lighter Regulatory Approach

President Donald Trump has taken a different stance on AI regulation, favoring a lighter touch and arguing that America must not slow AI development because "whoever wins with AI wins"2

2

. Trump dismissed existential AI threats as a hoax but announced plans to establish an AI Force similar to the Space Force and appoint an AI czar, stating that only high IQ individuals should apply, though he provided no further structural details1

1

. A June executive order indicated the administration would avoid overly burdensome regulation while focusing on cybersecurity and protecting critical infrastructure2

2

.

Growing Bipartisan Concerns Over Existential Risks

Bipartisan concerns about AI safety have intensified following warnings from a former Anthropic researcher about existential risks to human life1

1

. Former President Obama criticized the White House adviser's suggestion that the market would self-regulate AI, pointing out that such an approach contradicts established regulatory practices for airlines, drug companies, and food companies1

1

. Obama supported voluntary industry restraint to slow AI advancement until comprehensive solutions emerge from the White House and Congress1

1

. California Governor Gavin Newsom has also ordered the state to advance creation of AI kill switches, criticizing Washington's failure to lead on AI oversight2

2

. The geopolitical urgency is further highlighted by Treasury Secretary Scott Bessent's proposal for a notification mechanism for AI incidents reaching national-security levels, seeking greater transparency between the US and China2

2

.