New Survey: Most Americans Say Tech Companies Should Be Allowed to Set AI Limits as Anthropic Rejects Pentagon Demands

WASHINGTON—Amid Anthropic's dispute with the Pentagon over the use of its AI systems, 67% of Americans say private technology companies have a responsibility to set limits on how their products are used, even when the government disagrees. A further 53% say companies should have the right to restrict their AI from applications such as domestic surveillance and autonomous weapons, according to a new survey commissioned by the Information Technology and Innovation Foundation (ITIF), the leading think tank for science and technology policy.

The survey, conducted by Morning Consult among a nationally representative sample of 1,976 U.S. adults, examines public attitudes toward AI in military operations, surveillance, and corporate responsibility surrounding the Pentagon–Anthropic dispute. The findings reveal strong bipartisan support for keeping humans in control of lethal decisions, restricting the use of AI for domestic surveillance, and allowing technology companies to set boundaries on how their AI is used.

Support is bipartisan for allowing companies to set guardrails on their AI systems, though it narrows significantly when the question becomes more specific. Democrats (73%) and Republicans (65%) agree that technology companies have a responsibility to set limits on their products. However, when asked whether companies should be allowed to restrict their AI from uses like domestic surveillance or autonomous weapons, Republican support drops to 43% while Democratic support remains largely steady at 58%.

Key findings include:

• 79% say a human should always make the final decision before any use of lethal force (Democrats: 81%; Republicans: 81%)
• Majorities are “very concerned” about certain AI risks across the board:
  • Systems making errors in high-stakes military situations with no human to intervene (Adults: 54%; Democrats: 61%; Republicans: 51%)
  • AI being used for mass surveillance of American citizens by the government (Adults: 53%; Democrats: 59%; Republicans: 48%)
  • Other nations using AI for military purposes against the U.S. (Adults: 53%; Democrats: 55%; Republicans: 55%)
• 71% agree the U.S. should research and develop AI-controlled weapons to understand the technology and defend against adversaries, even if never deployed (Democrats: 72%; Republicans: 79%)
• 46% say the government should only use AI surveillance on specific targets with a court-issued warrant (Democrats: 45%; Republicans: 51%)
• 70% agree that using AI to monitor Americans without a warrant violates the Fourth Amendment (Democrats: 74%; Republicans: 71%)
• 45% oppose using emergency laws to compel AI company compliance (Democrats: 57%; Republicans: 29%)

On the dispute itself, 50% of Americans view penalizing Anthropic as government overreach that sets a dangerous precedent, versus 35% who say it is necessary for national security.

"Americans across party lines believe technology companies have a role in setting guardrails on how their AI systems are used,” said ITIF Vice President Daniel Castro. “At the same time, decisions about military AI cannot be settled through ad hoc standoffs between the Pentagon and individual firms. If certain AI capabilities are deemed essential for national defense, those expectations should be debated openly and written into law. Clear statutory boundaries give the military certainty, companies predictability, and the public confidence that democratic safeguards remain in place.”

View the findings here.

Contact: Nicole Hinojosa, [email protected]