{"id":26692,"date":"2026-03-05T19:55:20","date_gmt":"2026-03-05T19:55:20","guid":{"rendered":"https:\/\/microvibenews.com\/?p=26692"},"modified":"2026-03-05T19:55:20","modified_gmt":"2026-03-05T19:55:20","slug":"the-anthropic-openai-feud-and-their-pentagon-dispute-expose-a-deeper-problem-with-ai-safety","status":"publish","type":"post","link":"https:\/\/microvibenews.com\/?p=26692","title":{"rendered":"The Anthropic\u2013OpenAI feud and their Pentagon dispute expose a deeper problem with AI safety"},"content":{"rendered":"

<\/p>\n

Welcome to Eye on AI, with AI reporter Sharon Goldman. In this edition: Trump has an AI data center problem ahead of the midterms\u2026Don\u2019t trust AI to file your taxes\u2026Anthropic\u2019s AI tool Claude is central to US campaign in Iran, amid a bitter feud.<\/em><\/p>\n

\n

The debate around AI safety often focuses on the technology itself\u2014how powerful models might become, or what risks they might pose. But the conflict this week involving Anthropic, OpenAI and the Pentagon points to a deeper problem: how much power over the future of AI is concentrated in the hands of a small number of corporate leaders and government officials deciding how these systems are built, deployed, and used.<\/p>\n

For years, critics of the industry have warned about the risk of \u201cindustrial capture\u201d\u2014a future in which the development of powerful AI systems is concentrated among a handful of companies working closely with governments, leaving the safety of those systems dependent on the incentives and rivalries of the people running them. In 2023, for example, researcher Yoshua Bengio said the potential for the AI sector to be controlled by a few companies was the \u201cnumber two problem\u201d behind the existential risks posed by the technology.<\/p>\n

So it\u2019s not particularly reassuring to read yesterday about the disdain Anthropic CEO Dario Amodei expressed towards OpenAI CEO Sam Altman in leaked memo Amodei wrote to employees on Friday. Amodei\u2019s angry missive, which was apparently sent over Anthropic\u2019s Slack to all its employees, came after OpenAI announced a deal to provide AI to the Pentagon and Secretary of War Pete Hegseth said he was declaring Anthropic a \u201csupply chain risk\u201d for failing to come to a similar deal.<\/p>\n

Amodei called OpenAI\u2019s messaging \u201cmendacious,\u201d \u201csafety theater,\u201d and \u201can example of who they really are,\u201d while describing many of Altman\u2019s comments as \u201cstraight up lies\u201d and \u201cgaslighting.\u201d<\/p>\n

Altman has taken his own public shots at Anthropic. He recently called one of the company\u2019s Super Bowl campaigns \u201cclearly dishonest\u201d and accused it of \u201cdoublespeak.\u201d And the rivalry has become visible in more symbolic ways as well: At a recent summit, Altman and Amodei went viral for refusing to hold hands for a group photo with Prime Minister Narendra Modi.\u00a0<\/p>\n

With the US government taking little action to regulate AI\u2014and international efforts on AI safety largely stalled\u2014the world has effectively been relying on self-regulation by the industry. Both OpenAI and Anthropic have publicly supported that paradigm and signed voluntary safety commitments. They have also collaborated at times to run independent safety evaluations of one another\u2019s models prior to those models being released.<\/p>\n

But when the leaders of the two most influential AI labs so obviously can\u2019t seem to get along, and the competition between them is so fierce, it raises an uncomfortable question: how much cooperation on safety can we realistically expect?<\/p>\n

The pressure of competition has already impacted both companies when it comes to AI safety.\u00a0 Anthropic recently revised its Responsible Scaling Policy to say it would no longer unilaterally hold back from developing a new model simply because it did not yet know how to make that model safe. And OpenAI has made its own adjustments, removing explicit bans on military and warfare uses from its policies in 2024, and shifting its focus from safety research to product development to the point that former superalignment lead Jan Leike (who left for Anthropic in mid-2024) wrote on X that at OpenAI \u201csafety culture and processes have taken a backseat to shiny products.\u201d<\/p>\n

The current safety approach assumes that companies and governments will ultimately act with restraint. But the future of AI safety may ultimately depend on how a small number of powerful players navigate the pressures of competition, geopolitics, and the occasional Silicon Valley soap opera.<\/p>\n

With that, here\u2019s more AI news.<\/p>\n

Sharon Goldman<\/strong>
sharon.goldman@fortune.com
@sharongoldman<\/p>\n

FORTUNE ON AI<\/h3>\n

Why Leopold Aschenbrenner\u2019s AI hedge fund is betting big on power companies and bitcoin miners to fuel the \u2018superintelligence\u2019 race<\/span> \u2013 by Sharon Goldman<\/span><\/p>\n

OpenAI sees Codex users spike to 1.6 million, positions coding tool as gateway to AI agents for business<\/span> \u2013 by Jeremy Kahn<\/span><\/p>\n

Korean startup wrtn is on track to pass $100M in annual recurring revenue, riding a loneliness epidemic-fueled boom in AI entertainment<\/span> \u2013 by Nicolas Gordon<\/span><\/p>\n

AI IN THE NEWS<\/h3>\n

Trump has an AI data center problem ahead of the midterms.<\/strong> CNBC and others reported that President Trump<\/span> is facing a growing political dilemma as the U.S. races to build energy-hungry AI data centers ahead of the 2026 midterms. The infrastructure needed to power the AI boom is driving concerns about rising electricity prices and strain on the grid, prompting backlash from voters and local communities. In response, major tech companies\u2014including OpenAI<\/span>, Microsoft<\/span>, Google<\/span>, Amazon<\/span>, Meta<\/span>, and Oracle<\/span>\u2014have pledged to cover the energy and infrastructure costs associated with their AI data centers so that consumers don\u2019t see higher utility bills. The voluntary agreement, promoted by the White House as a way to ease voter concerns, reflects a broader tension: policymakers want the economic and geopolitical advantages of rapid AI expansion, but the enormous electricity demands of the technology are creating political and environmental pressures that are becoming harder to ignore.<\/p>\n

\n
\n
\n
\n
\n
\n
\n
\n

Don’t trust AI to file your taxes.<\/strong> In results that should surprise no one, a test by The New York Times<\/em> found that AI is no match for the US tax code, highlighting an important limitation of today\u2019s AI chatbots: they still struggle with tasks that require precise, multi-step reasoning. To assess the technology\u2019s ability to file a federal income tax return, the paper tested four AI chatbots \u2014 Google\u2019s Gemini, OpenAI\u2019s ChatGPT, Anthropic\u2019s Claude and xAI\u2019s Grok \u2014 to see how well they fared with eight fictional tax situations.\u00a0They struggled, hard, miscalculating the refund or amount owed to the Internal Revenue Service by an average of more than $2,000. Even when provided with all the necessary materials, including all the forms they needed to fill out, the chatbots whiffed on some calculations. The problem reflects a fundamental limitation of large language models: they are designed to predict likely words rather than precisely track complex, interconnected information, making them strong at writing and summarization but weaker at procedural tasks like tax filing. Experts say the systems may improve with additional reasoning tools and verification layers, but for now they work best as assistants rather than replacements\u2014another reminder that even as AI reshapes industries from coding to medicine, some seemingly simpler tasks remain surprisingly difficult.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n

\n

Anthropic\u2019s AI tool Claude is central to US campaign in Iran, amid a bitter feud.<\/strong> A new report from The Washington Post<\/em> highlights how quickly AI has moved from experimentation to the battlefield. According to the paper, the US military used an AI-enabled targeting system called Maven Smart System\u2014built by Palantir<\/span><\/span> and incorporating Anthropic<\/span><\/span>\u2019s Claude model\u2014to help identify and prioritize targets during recent U.S. operations in Iran, accelerating what once took weeks of military planning into near-real-time decision making. Yet the deployment comes amid a bitter dispute between Anthropic and the Pentagon over limits on how its technology can be used in warfare, including concerns about autonomous weapons and mass surveillance. The episode underscores both the growing strategic importance of frontier AI systems and the tension between government demand for rapid deployment and companies\u2019 attempts to set safety boundaries.\u00a0<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/article>\n<\/div>\n

EYE ON AI NUMBERS<\/h3>\n

$25 billion<\/h2>\n

That\u2019s how much annualized revenue OpenAI was generating as of the end of last month, according to reporting by The Information<\/em>\u2014a 17% jump from the $21.4 billion annualized run rate it had at the end of the year, according to two people familiar with the figures.<\/p>\n

OpenAI still brings in more revenue than its closest rival, Anthropic, though the gap is quickly narrowing. Anthropic\u2019s annualized revenue recently topped $19 billion, nearly triple what it was at the end of last year and up 36% in just the past two weeks.<\/p>\n

OpenAI calculates annualized revenue by multiplying the previous four weeks of revenue by 12. One source said that if the company instead extrapolated from revenue spikes in the most recent week alone, its annualized run rate would be closer to $30 billion.<\/p>\n

Anthropic\u2019s rapid growth has been fueled in part by strong demand for its coding-focused AI models, which have helped the company quickly narrow the revenue gap with OpenAI. As recently as 2025, OpenAI was generating roughly three times as much revenue as Anthropic.<\/p>\n

AI CALENDAR<\/h3>\n

March 2-5:\u00a0<\/strong>Mobile World Congress, Barcelona, Spain.<\/p>\n

March 12-18:\u00a0<\/strong>South by Southwest, Austin, Texas.<\/p>\n

March 16-19:\u00a0<\/strong>Nvidia GTC, San Jose, Calif.<\/p>\n

April 6-9:<\/strong> HumanX, San Francisco.\u00a0<\/p>\n<\/div>\n

#AnthropicOpenAI #feud #Pentagon #dispute #expose #deeper #problem #safety<\/p>\n","protected":false},"excerpt":{"rendered":"

Welcome to Eye on AI, with AI … <\/p>\n","protected":false},"author":1,"featured_media":26693,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":[],"categories":[2],"tags":[12442,4298,704,15183,4296,516,7262,13177,3432,7261,2936,15184,4076,703,10304,13944,1131,214,2787,15182,4297],"_links":{"self":[{"href":"https:\/\/microvibenews.com\/index.php?rest_route=\/wp\/v2\/posts\/26692"}],"collection":[{"href":"https:\/\/microvibenews.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/microvibenews.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/microvibenews.com\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/microvibenews.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=26692"}],"version-history":[{"count":0,"href":"https:\/\/microvibenews.com\/index.php?rest_route=\/wp\/v2\/posts\/26692\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/microvibenews.com\/index.php?rest_route=\/wp\/v2\/media\/26693"}],"wp:attachment":[{"href":"https:\/\/microvibenews.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=26692"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/microvibenews.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=26692"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/microvibenews.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=26692"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}