House Committees Investigate DoorDash's Use of Chinese AI

4 hours ago

House committees are investigating the national security, cybersecurity, and economic security risks of U.S. companies using Chinese open-weight AI models, focusing on DoorDash.
Open-weight AI models lack guardrails, making them vulnerable to misuse, such as code injection or attacks on AI systems.
The investigation aims to gather facts to inform a federal approach that reduces reliance on Chinese models while promoting American open-weight alternatives with appropriate safeguards.
Previous letters were sent to Anysphere and Airbnb in April 2026, and DoorDash is now requested to provide similar information.
Chinese efforts to close the AI innovation gap through model distillation and intellectual property theft heighten the national security threat.

US lawmakers are investigating DoorDash for using Moonshot AI's Kimi K2.6 model, a Chinese-developed AI.
The investigation reflects growing tensions as US companies increasingly adopt Chinese AI models to cut costs.
Lawmakers request an in-person briefing from DoorDash staff by August 21 on AI infrastructure and security.

OpenRouter aggregates AI models from various providers, enabling developers to choose efficient options.
Chinese models, particularly DeepSeek, are rising in popularity on OpenRouter, reflecting a shift in US companies' AI access.
Flo Crivello notes a potential Chinese cutoff would impact firms but not be catastrophic, with US companies needing 3-6 months to match Chinese open-weight models.

Model distillation is increasingly seen as a national security concern, with Chinese AI firms like Zhipu accused of extracting capabilities from U.S. frontier models.
Mira Murati's Thinking Machines built its first foundation model using Chinese models such as DeepSeek-V3 and Moonshot AI's Kimi K2.5, highlighting reciprocal innovation.
Distillation is a common machine learning technique, not new or inherently nefarious, but now politically charged in the context of U.S.-China AI competition.

The most serious test to date for open source AI viability is happening now due to policy actions that could marginalize open models.
White House discussions may lead to new executive orders impacting open models, potentially starting with Chinese-origin models and government uses.
Open models lack a central economic champion to defend against regulatory actions, unlike closed models with strong lobbying.

Anthropic and its CEO, Dario Amodei, do not advocate for a ban on open-weights models, viewing them as a public good when lacking dangerous capabilities.
The primary concern is authoritarian governments, particularly China, using powerful AI for military or repressive purposes, which is unrelated to open-weights access by US businesses.
A secondary concern is the misuse of powerful AI models for cyber or biological attacks; open-weights increase risk due to difficulty in applying guardrails, but banning US business use does not address this.

US lawmakers are investigating DoorDash for using Moonshot AI's Kimi K2.6 model, a Chinese-developed AI.
The investigation reflects growing tensions as US companies increasingly adopt Chinese AI models to cut costs.
Lawmakers request an in-person briefing from DoorDash staff by August 21 on AI infrastructure and security.

OpenRouter aggregates AI models from various providers, enabling developers to choose efficient options.
Chinese models, particularly DeepSeek, are rising in popularity on OpenRouter, reflecting a shift in US companies' AI access.
Flo Crivello notes a potential Chinese cutoff would impact firms but not be catastrophic, with US companies needing 3-6 months to match Chinese open-weight models.

Model distillation is increasingly seen as a national security concern, with Chinese AI firms like Zhipu accused of extracting capabilities from U.S. frontier models.
Mira Murati's Thinking Machines built its first foundation model using Chinese models such as DeepSeek-V3 and Moonshot AI's Kimi K2.5, highlighting reciprocal innovation.
Distillation is a common machine learning technique, not new or inherently nefarious, but now politically charged in the context of U.S.-China AI competition.

The most serious test to date for open source AI viability is happening now due to policy actions that could marginalize open models.
White House discussions may lead to new executive orders impacting open models, potentially starting with Chinese-origin models and government uses.
Open models lack a central economic champion to defend against regulatory actions, unlike closed models with strong lobbying.

Anthropic and its CEO, Dario Amodei, do not advocate for a ban on open-weights models, viewing them as a public good when lacking dangerous capabilities.
The primary concern is authoritarian governments, particularly China, using powerful AI for military or repressive purposes, which is unrelated to open-weights access by US businesses.
A secondary concern is the misuse of powerful AI models for cyber or biological attacks; open-weights increase risk due to difficulty in applying guardrails, but banning US business use does not address this.

Hasty Briefsbeta