Marco Rubio tells US diplomats to downplay tech kill switch fears after Anthropic AI ban fallout

EXCLUSIVE: Marco Rubio tells diplomats to play down talk of American tech 'kill switch'

WASHINGTON, July 22 (Reuters) - U.S. Secretary of State Marco Rubio has asked diplomats to push back against talk of a "kill switch" in American technology products following the White House's short-lived decision to keep foreigners from America's most advanced AI models, according to a recent cable reviewed by Reuters. The talking points, which were circulated worldwide, show how U.S. diplomats are trying to deal with international backlash from the Trump administration's efforts to control how and to whom American AI companies release their models. The State Department declined to comment and the White House did not return a request for comment. The Trump administration's June 12 move to block non-U.S. residents from Anthropic's most advanced models, Mythos and Fable, on national security grounds prompted the AI company to abruptly suspend access to the models globally. Asian tech executives seized on the disruption to tout their own alternatives and European lawmakers redoubled their calls for digital independence from ⁠the United States. Although the ban was lifted later the same month, the fallout lingers. Combined with Trump's June 2 executive order, which asked AI companies to voluntarily submit their models to 30 days of cybersecurity testing ahead of their release, it signaled a new appetite from the U.S. government to police AI companies' products -- and potentially yank them even from the hands of allies at a moment's notice. European lawmaker Christophe Grudler said, opens new tab the episode showed that "the U.S. holds a real 'kill-switch' over essential technologies and that they are more than willing to use it." Aura Salla, a member of the European People's Party, the European Parliament's biggest political group, said, opens new tab that the continent "cannot keep building its tech stack on access that can be switched off overnight by a foreign government." The State Department cable, dated July 16, did not mention Anthropic by name or refer specifically to last month's ban or Trump's executive order, but it provided American diplomats with talking points aimed at countering the arguments that have sprung up in their wake. "Pausing narrow uses or requiring a 30-day testing window prior to the release of a highly potent new ⁠technology is not a 'Kill Switch'," the cable said. "There is no government 'magic button.' This narrative is exaggerated and doesn't capture the nuances of U.S. technology policy." AMERICAN AI SALES PITCH Rubio's cable pushed diplomats to fight "so-called 'digital sovereignty'" initiatives, which it defined as efforts to restrict American tech firms' access to foreign markets, subject them to localization requirements, charge them "network usage fees," or force them to follow local rules around issues such as content moderation. Rubio had already instructed American diplomats to oppose similar data sovereignty measures earlier this year, Reuters previously reported. The cable also discussed how to counter "AI sovereignty" arguments, instructing American diplomats to ⁠advertise American AI products as the best tools available and describe efforts to build rival AI systems from the ground up as a waste of time and resources. "American AI companies can build large, independent AI infrastructure, with secure and robust supply chains that minimize backdoor risk," the cable said. "They build it. It's yours." Edward Fishman, director of the Maurice R. Greenberg Center for Geoeconomics at the Council on ⁠Foreign Relations, said he sympathized with the safety considerations that prompted the U.S. government to first impose restrictions on Anthropic's models, which can turbocharge complex hacking operations. But Fishman said that the Trump administration's habit of wielding coercive economic tools such as tariffs and sanctions against friend and foe alike made the State Department's pitch a tough sell. European partners in particular were ⁠concerned that if their businesses became "ultra reliant on U.S. frontier AI models like those made by Anthropic and OpenAI, that the U.S. could use that as an economic weapon against them down the road," he said. Salla, the European lawmaker, told Reuters that the Trump administration's repeated threats against countries like Denmark had generated an atmosphere of mistrust that no amount of rhetoric could dispel. "They have not shown us any indication that they are reliable partners," she said. Reporting by Raphael Satter, editing by Deepa Babington Our Standards: The Thomson Reuters Trust Principles., opens new tab * Suggested Topics: * Artificial Intelligence * Data Privacy * Public Policy Raphael Satter Thomson Reuters Reporter covering cybersecurity, surveillance, and disinformation for Reuters. Work has included investigations into state-sponsored espionage, deepfake-driven propaganda, and mercenary hacking.

Marco Rubio tells US diplomats to play down fears of a tech 'kill switch'

US Secretary of State Marco Rubio has told American diplomats to push back against a phrase that has spread through allied capitals since the summer: the tech "kill switch." In a cable dated July 16 and reviewed by Reuters, Rubio instructed envoys to counter the idea that Washington can remotely disable American technology abroad, a fear that hardened after the government briefly blocked foreign access to Anthropic's top models in June and revived European talk of digital sovereignty. "Pausing narrow uses or requiring a 30-day testing window prior to release of a highly potent new technology is not a 'Kill Switch,'" the cable reads, according to Reuters' account of the document. "There is no government 'magic button.'" It calls the narrative exaggerated and tells diplomats it fails to capture the nuances of US technology policy. The anxiety has a specific origin. On June 12, the Trump administration barred non-US residents from Anthropic's most advanced models, Mythos and Fable, on national security grounds, then reversed course within the same month. The whiplash left officials across Europe and Asia wondering how much of their AI stack sat behind a switch they did not control, and it sent technology executives in several capitals scrambling to promote home-grown alternatives. That block followed a June 2 executive order requiring AI developers to submit new models for 30 days of cybersecurity testing before release. For governments running public services on top of American models, the sequence looked less like routine safety review and more like leverage. Brussels held talks with Washington over restoring access, and the shutdown became a live case study in dependence. Canadian Prime Minister Mark Carney went as far as comparing it to the 2008 financial crisis, warning of a systemic AI "model risk" that no single government could contain. European lawmakers said as much in public. Christophe Grudler and Aura Salla, both of the centre-right European People's Party, warned that the United States held "a real 'kill-switch' over essential technologies" that could be "switched off overnight by a foreign government." Edward Fishman of the Council on Foreign Relations told Reuters that allies increasingly fear their reliance on American artificial intelligence leaves them exposed to economic coercion. EU finance ministers had already pressed for guaranteed access to Anthropic's systems, and Austria floated the idea of hosting the company on European soil. Not everyone accepted the framing. The International Association of Privacy Professionals argued the Anthropic episode was a governance problem rather than a kill switch, a security challenge dressed up as a sovereignty crisis. Rubio's cable leans on that reading and pushes it further. The document does more than quibble over vocabulary. It instructs diplomats to discourage "so-called 'digital sovereignty'" initiatives that would wall off American products, and to promote US AI as the superior option wherever foreign governments weigh home-grown alternatives. The message, in effect, is that dependence is not a risk to be managed but a market to be defended. The phrase itself did not appear from nowhere. For more than a year, US lawmakers have floated the Chip Security Act, which would mandate location tracking, and in some drafts a remote kill switch, on exported Nvidia processors to keep them out of China. A separate Remote Access Security Act sought to extend export controls into the cloud, closing the loophole that let restricted chips be rented offshore. Nvidia has repeatedly denied that its chips carry backdoors or kill switches and urged Congress against location-verification mandates. Washington's own chip curbs have tightened around Chinese firms in parallel. Whether a testing window and a temporary export block amount to a switch is partly a question of language and partly one of trust. Marco Rubio is betting his diplomats can win the argument over the word. The harder problem, and the one the cable does not resolve, is that allies now know the mechanism exists, whatever anyone agrees to call it.

Exclusive-Marco Rubio Tells Diplomats to Play Down Talk of American Tech 'Kill Switch'

WASHINGTON, July 22 (Reuters) - U.S. Secretary of State Marco Rubio has asked diplomats to ⁠push ⁠back against talk of a "kill switch" in American technology ⁠products following the White House's short-lived decision to keep foreigners from America's most advanced AI models, according to a recent cable reviewed by Reuters. The talking points, which were circulated worldwide, show how U.S. diplomats are trying to deal with international backlash from the Trump administration's efforts to control how and to whom American AI companies release their models. The State Department declined to comment and the White House did not return a request for comment. The Trump administration's June 12 move to block non-U.S. residents from Anthropic's most advanced models, Mythos and Fable, ⁠on national security grounds ⁠prompted the AI company to abruptly suspend access to the models globally. Asian tech executives seized on the disruption to tout their own alternatives and European lawmakers redoubled their calls for digital independence from the United States. Although the ban was lifted later the same month, the fallout lingers. Combined with Trump's June 2 executive order, which asked AI companies to voluntarily submit their models to 30 days of cybersecurity testing ahead of their release, it signaled a new appetite from the U.S. government to police AI companies' products -- and potentially yank them even from the hands of allies at a moment's notice. European lawmaker Christophe Grudler said the episode showed that "the ⁠U.S. ⁠holds a real 'kill-switch' over essential technologies and ⁠that they are more than willing to use it." Aura Salla, a member of the European People's Party, the European Parliament's biggest political group, said that the continent "cannot keep building its tech stack on access that can be switched ⁠off overnight by a foreign government." The State Department cable, dated July 16, did not mention Anthropic by name or refer specifically to last month's ban or Trump's executive order, but it provided American diplomats with talking points aimed at countering the arguments that have sprung up in their wake. "Pausing narrow uses or requiring a 30-day testing window prior to the release of a highly potent new technology is not a 'Kill Switch'," the cable said. "There is no government 'magic button.' This narrative is exaggerated and doesn't capture the nuances of U.S. technology policy." AMERICAN ⁠AI SALES PITCH Rubio's cable pushed diplomats to fight "so-called 'digital sovereignty'" initiatives, which it defined as efforts to restrict American tech firms' access to ⁠foreign markets, subject them to localization requirements, charge them "network usage fees," or force them to follow local rules around issues such as content moderation. Rubio had already instructed American diplomats to oppose similar data sovereignty measures earlier this year, Reuters previously reported. The cable also discussed how to counter "AI sovereignty" arguments, instructing American diplomats to advertise American AI products as the best tools available and describe efforts to build rival AI systems from the ground up as a waste of time and resources. "American AI companies can build large, independent AI infrastructure, with secure and robust supply chains that minimize backdoor risk," the cable said. "They build it. It's yours." Edward Fishman, director of the Maurice R. Greenberg Center for Geoeconomics at the Council on Foreign Relations, said he sympathized with the safety considerations that prompted the U.S. government to first impose restrictions on Anthropic's models, which can turbocharge complex hacking operations. But Fishman said that the Trump administration's ⁠habit of wielding coercive economic tools such as tariffs and sanctions against friend and foe alike made the State Department's pitch a tough sell. European partners in particular were concerned that if their businesses became "ultra reliant on U.S. frontier AI models like those made by Anthropic and OpenAI, that the U.S. could use that as an economic weapon against them down the road," he said. Salla, the European lawmaker, told Reuters that the Trump administration's repeated threats against countries like Denmark had generated an atmosphere of mistrust that no amount of rhetoric could dispel. "They have not shown us any indication that they are reliable partners," she said. (Reporting by Raphael Satter, editing by Deepa Babington)

Rubio tells diplomats to push back on tech 'kill switch' talk after attempt to block AI companies

US Secretary of State Marco Rubio has asked diplomats to push back against talk of a "kill switch" in American technology products following the White House's short-lived decision to keep foreigners from America's most advanced AI models, according to a recent cable reviewed by Reuters. The talking points, which were circulated worldwide, show how US diplomats are trying to deal with the international backlash from the Trump administration's efforts to control how and to whom American AI companies release their models. The White House referred questions to the State Department, which in turn pointed Reuters to an editorial from US Under Secretary of State for Economic Affairs Jacob Helberg that decried "autarkist measures" which aimed to push American technology aside in favor of native AI solutions, a trend he described as "backward and counterproductive." Rubio's missive follows the Trump administration's June 12 move to block foreign nationals from Anthropic's most advanced models, Mythos and Fable, on national security grounds, prompting the AI company to abruptly suspend access to the models globally. Asian tech executives seized on the disruption to tout their own alternatives and European lawmakers redoubled their calls for digital independence from the United States. Although the ban was lifted later the same month, the fallout lingers. Combined with Trump's June 2 executive order, which asked AI companies to voluntarily submit their models to 30 days of cybersecurity testing ahead of their release, it signaled a new appetite from the US government to police AI companies' products - and potentially yank them even from the hands of allies at a moment's notice. European lawmaker Christophe Grudler said the episode showed that "the US holds a real 'kill-switch' over essential technologies and that they are more than willing to use it." Aura Salla, a member of the European People's Party, the European Parliament's biggest political group, said that the continent "cannot keep building its tech stack on access that can be switched off overnight by a foreign government." Rubio tells diplomats how to combat 'AI sovereignty' arguments The State Department cable, dated July 16, did not mention Anthropic by name or refer specifically to last month's ban or Trump's executive order, but it provided American diplomats with talking points aimed at countering the arguments that have sprung up in their wake. "Pausing narrow uses or requiring a 30-day testing window prior to the release of a highly potent new technology is not a 'Kill Switch'," the cable said. "There is no government 'magic button.' This narrative is exaggerated and doesn't capture the nuances of US technology policy." Rubio's cable pushed diplomats to fight "so-called 'digital sovereignty'" initiatives, which it defined as efforts to restrict American tech firms' access to foreign markets, subject them to localization requirements, charge them "network usage fees," or force them to follow local rules around issues such as content moderation. Rubio had already instructed American diplomats to oppose similar data sovereignty measures earlier this year, Reuters previously reported. The cable also discussed how to counter "AI sovereignty" arguments, instructing American diplomats to advertise American AI products as the best tools available and describe efforts to build rival AI systems from the ground up as a waste of time and resources. "American AI companies can build large, independent AI infrastructure, with secure and robust supply chains that minimize backdoor risk," the cable said. "They build it. It's yours." Europe concerned over Trump administration's threats Edward Fishman, director of the Maurice R. Greenberg Center for Geoeconomics at the Council on Foreign Relations, said he sympathized with the safety considerations that prompted the US government to first impose restrictions on Anthropic's models, which can turbocharge complex hacking operations. But Fishman said the Trump administration's habit of wielding coercive economic tools such as tariffs and sanctions against friend and foe alike made the State Department's pitch a tough sell. European partners in particular were concerned that if their businesses became "ultra reliant on US frontier AI models like those made by Anthropic and OpenAI, that the US could use that as an economic weapon against them down the road," he said. Salla, the European lawmaker, told Reuters that the Trump administration's repeated threats against countries like Denmark had generated an atmosphere of mistrust that no amount of rhetoric could dispel. "They have not shown us any indication that they are reliable partners," she said.

Marco Rubio tells diplomats to play down talk of American tech 'kill switch': Reuters exclusive

WASHINGTON -- U.S. Secretary of State Marco Rubio has asked diplomats to push back against talk of a "kill switch" in American technology products following the White House's short-lived decision to keep foreigners from America's most advanced AI models, according to a recent cable reviewed by Reuters. The talking points, which were circulated worldwide, show how U.S. diplomats are trying to deal with international backlash from the Trump administration's efforts to control how and to whom American AI companies release their models. The State Department declined to comment and the White House did not return a request for comment. The Trump administration's June 12 move to block non-U.S. residents from Anthropic's most advanced models, Mythos and Fable, on national security grounds prompted the AI company to abruptly suspend access to the models globally. Asian tech executives seized on the disruption to tout their own alternatives and European lawmakers redoubled their calls for digital independence from the United States. Although the ban was lifted later the same month, the fallout lingers. Combined with Trump's June 2 executive order, which asked AI companies to voluntarily submit their models to 30 days of cybersecurity testing ahead of their release, it signaled a new appetite from the U.S. government to police AI companies' products -- and potentially yank them even from the hands of allies at a moment's notice. European lawmaker Christophe Grudler said the episode showed that "the U.S. holds a real 'kill-switch' over essential technologies and that they are more than willing to use it." Aura Salla, a member of the European People's Party, the European Parliament's biggest political group, said that the continent "cannot keep building its tech stack on access that can be switched off overnight by a foreign government." The State Department cable, dated July 16, did not mention Anthropic by name or refer specifically to last month's ban or Trump's executive order, but it provided American diplomats with talking points aimed at countering the arguments that have sprung up in their wake. "Pausing narrow uses or requiring a 30-day testing window prior to the release of a highly potent new technology is not a 'Kill Switch'," the cable said. "There is no government 'magic button.' This narrative is exaggerated and doesn't capture the nuances of U.S. technology policy." American AI sales pitch Rubio's cable pushed diplomats to fight "so-called 'digital sovereignty'" initiatives, which it defined as efforts to restrict American tech firms' access to foreign markets, subject them to localization requirements, charge them "network usage fees," or force them to follow local rules around issues such as content moderation. Rubio had already instructed American diplomats to oppose similar data sovereignty measures earlier this year, Reuters previously reported. The cable also discussed how to counter "AI sovereignty" arguments, instructing American diplomats to advertise American AI products as the best tools available and describe efforts to build rival AI systems from the ground up as a waste of time and resources. "American AI companies can build large, independent AI infrastructure, with secure and robust supply chains that minimize backdoor risk," the cable said. "They build it. It's yours." Edward Fishman, director of the Maurice R. Greenberg Center for Geoeconomics at the Council on Foreign Relations, said he sympathized with the safety considerations that prompted the U.S. government to first impose restrictions on Anthropic's models, which can turbocharge complex hacking operations. But Fishman said that the Trump administration's habit of wielding coercive economic tools such as tariffs and sanctions against friend and foe alike made the State Department's pitch a tough sell. European partners in particular were concerned that if their businesses became "ultra reliant on U.S. frontier AI models like those made by Anthropic and OpenAI, that the U.S. could use that as an economic weapon against them down the road," he said. Salla, the European lawmaker, told Reuters that the Trump administration's repeated threats against countries like Denmark had generated an atmosphere of mistrust that no amount of rhetoric could dispel. "They have not shown us any indication that they are reliable partners," she said.

Marco Rubio tells diplomats to play down talk of American tech 'kill switch'

WASHINGTON, July 22 (Reuters) - U.S. Secretary of State Marco Rubio has asked diplomats to push back against talk of a "kill switch" in American technology products following the White House's short-lived decision to keep foreigners from America's most advanced AI models, according to a recent cable reviewed by Reuters. The talking points, which were circulated worldwide, show how U.S. diplomats are trying to deal with international backlash from the Trump administration's efforts to control how and to whom American AI companies release their models. The State Department declined to comment and the White House did not return a request for comment. The Trump administration's June 12 move to block non-U.S. residents from Anthropic's most advanced models, Mythos and Fable, on national security grounds prompted the AI company to abruptly suspend access to the models globally. Asian tech executives seized on the disruption to tout their own alternatives and European lawmakers redoubled their calls for digital independence from the United States. Although the ban was lifted later the same month, the fallout lingers. Combined with Trump's June 2 executive order, which asked AI companies to voluntarily submit their models to 30 days of cybersecurity testing ahead of their release, it signaled a new appetite from the U.S. government to police AI companies' products -- and potentially yank them even from the hands of allies at a moment's notice. European lawmaker Christophe Grudler said the episode showed that "the U.S. holds a real 'kill-switch' over essential technologies and that they are more than willing to use it." Aura Salla, a member of the European People's Party, the European Parliament's biggest political group, said that the continent "cannot keep building its tech stack on access that can be switched off overnight by a foreign government." The State Department cable, dated July 16, did not mention Anthropic by name or refer specifically to last month's ban or Trump's executive order, but it provided American diplomats with talking points aimed at countering the arguments that have sprung up in their wake. "Pausing narrow uses or requiring a 30-day testing window prior to the release of a highly potent new technology is not a 'Kill Switch'," the cable said. "There is no government 'magic button.' This narrative is exaggerated and doesn't capture the nuances of U.S. technology policy." AMERICAN AI SALES PITCH Rubio's cable pushed diplomats to fight "so-called 'digital sovereignty'" initiatives, which it defined as efforts to restrict American tech firms' access to foreign markets, subject them to localization requirements, charge them "network usage fees," or force them to follow local rules around issues such as content moderation. Rubio had already instructed American diplomats to oppose similar data sovereignty measures earlier this year, Reuters previously reported. The cable also discussed how to counter "AI sovereignty" arguments, instructing American diplomats to advertise American AI products as the best tools available and describe efforts to build rival AI systems from the ground up as a waste of time and resources. "American AI companies can build large, independent AI infrastructure, with secure and robust supply chains that minimize backdoor risk," the cable said. "They build it. It's yours." Edward Fishman, director of the Maurice R. Greenberg Center for Geoeconomics at the Council on Foreign Relations, said he sympathized with the safety considerations that prompted the U.S. government to first impose restrictions on Anthropic's models, which can turbocharge complex hacking operations. But Fishman said that the Trump administration's habit of wielding coercive economic tools such as tariffs and sanctions against friend and foe alike made the State Department's pitch a tough sell. European partners in particular were concerned that if their businesses became "ultra reliant on U.S. frontier AI models like those made by Anthropic and OpenAI, that the U.S. could use that as an economic weapon against them down the road," he said. Salla, the European lawmaker, told Reuters that the Trump administration's repeated threats against countries like Denmark had generated an atmosphere of mistrust that no amount of rhetoric could dispel. "They have not shown us any indication that they are reliable partners," she said. (Reporting by Raphael Satter, editing by Deepa Babington)