A federal appeals court in Washington ruled that the Pentagon acted legally when it labeled Anthropic a security risk. The three-judge panel upheld the blacklist, though one judge dissented. The court found the Pentagon had sufficient evidence for its February decision and that Anthropic's First Amendment rights were not violated. The ruling also established that the US can blacklist a company for withholding specific AI features even without proving malicious intent, focusing on Anthropic's refusal to enable certain Claude capabilities.
Key Takeaways
- A federal appeals court upheld the Pentagon's blacklisting of Anthropic over national security concerns tied to Claude features
- The 2-1 ruling allows companies to be blacklisted for withholding AI features without proving malicious intent
- Anthropic researcher Jacob Coxon resigned over safety concerns and worked with PR firm DEY, raising questions about the AI safety debate
- A customizable router saves enterprises between $3.3 million and $5.0 million per year by cutting AI coding costs by up to 21%
- USPTO appointed Jonathan Spencer, a former Waymo engineer, as its new Chief AI Officer
- Meta's Muse AI agent drove a surge in Arm Holdings stock due to CPU demand for AI inference workloads
- Senator John Curtis cosponsored the AI Whistleblower Protection Act alongside senators from both parties
- A Doomsday Clock expert warned that AI poses an extremely pressing threat requiring immediate action
- A New York state audit found AI oversight gaps at SUNY campuses, with new policies required by end of 2026
- Sabina Matos accused Aaron Guckian of using AI to copy her policy idea, with tests showing 100% AI-generated content
Court rules Pentagon blacklisting of Anthropic was legal
A federal appeals court in Washington ruled that the Pentagon acted legally when it labeled Anthropic a security risk. The court said the Pentagon had enough evidence to support its decision from February. The ruling also said Anthropic's First Amendment rights were not violated. One judge dissented from the three-judge panel's decision.
Federal appeals court upholds Pentagon blacklist of Anthropic
A federal appeals court in Washington upheld the Pentagon's decision to blacklist Anthropic. The AI company was labeled a risk to national security. This type of label was once used mainly for foreign companies suspected of spying. Anthropic makes the Claude chatbot and is one of the country's top tech firms.
Court approves blacklisting Anthropic over Claude features
A US appeals court approved the Trump administration's blacklisting of Anthropic. Judges ruled the US could blacklist the company for withholding certain AI features even without malicious intent. The 2-1 ruling came from the US Court of Appeals for the District of Columbia Circuit. The decision focused on Anthropic's refusal to enable specific Claude features.
USPTO names Jonathan Spencer as Chief AI Officer
The US Patent and Trademark Office appointed Jonathan Spencer as chief AI officer. Spencer is a former software engineer at Waymo who spent four years developing AI models for autonomous vehicles. He will lead the agency's AI efforts and help deploy AI systems across patent and trademark operations. Spencer holds a doctorate from Princeton and has co-founded several AI startups.
Arm Holdings stock rises on AI CPU demand surge
Arm Holdings stock gained on September 23 2026 during a broader semiconductor rally. Investors focused on AI-driven CPU demand after Meta Platforms reported rapid uptake of its Muse AI agent. Arm's CPU designs were highlighted as a key foundation for handling Muse-related AI inference workloads. The surge reinforced Arm's importance in AI infrastructure.
Custom router cuts AI coding costs by up to 21 percent
Researchers built a customizable router that controls how enterprises use AI coding agents. The router picks cheaper models and avoids rebuilding prompt caches during long tasks. Tests on 10,000 simulated seats showed it recovers 14 to 21 percent of model spending. At current Anthropic prices, this saves between $3.3M and $5.0M per year.
Matos says Guckian used AI to copy her policy idea
Sabina Matos accused Aaron Guckian of using AI to copy her business succession plan. Guckian released a plan helping retiring doctors and business owners pass operations to employees. A Matos campaign test found Guckian's news release was 100 percent AI generated. Guckian's campaign did not deny AI use but criticized Matos for her national political travel.
Anthropic quitter Jacob Coxon worked with PR firm DEY
Jacob Coxon, an Anthropic researcher, resigned over AI safety concerns. A Pirate Wires report found he worked with PR firm DEY around his announcement. DEY represents figures like Noam Chomsky and AI safety voices Eliezer Yudkowsky and Timnit Gebru. The report raises questions about who shaped the sudden AI safety debate.
State audit finds AI oversight gaps at SUNY campuses
A New York state audit found gaps in AI oversight at SUNY campuses. Auditors reviewed Onondaga Community College and Upstate Medical University. OCC lacked a formal AI inventory and risk plan, though it plans to rebuild its governance. Upstate had an unapproved AI policy but now has a new framework. SUNY requires all campuses to update AI policies by the end of 2026.
Options traders can play Meta's AI driven stock breakout
Mike Khouw of YieldMaxETFs explained how options traders can respond to Meta's AI advances. The strategy focuses on Meta's hardware and AI developments. The video breaks down what these changes mean for traders watching the stock.
Curtis Cosponsors Bill to Protect AI Whistleblowers
U.S. Senator John Curtis (R-UT) cosponsored the AI Whistleblower Protection Act. The bill would protect AI workers who report legal violations, security vulnerabilities, or public safety risks. It would prevent employers from using nondisclosure agreements to silence these workers. The Department of Labor would handle complaints without creating a new agency. Senators Chris Coons, Marsha Blackburn, Amy Klobuchar, Josh Hawley, and Brian Schatz also cosponsored the legislation.
Doomsday Clock Expert Warns AI Is an Extremely Pressing Threat
A Doomsday Clock expert warned that artificial intelligence poses an extremely pressing threat. The expert said the danger is real and not just hypothetical. The threat involves the potential misuse of AI and its possible consequences. The expert stressed that action is needed now to prevent a catastrophe. Other experts and organizations have also raised alarms about AI causing widespread harm.
Sources
- Anthropic’s Blacklisting by the Pentagon Was Legal, Federal Judges Rule
- Federal appeals court rules Pentagon can blacklist Anthropic
- Court rules Trump can blacklist Anthropic for refusing to enable Claude features
- USPTO Appoints Jonathan Spencer as Chief AI Officer
- Arm Holdings (ARM) Stock Gains On AI CPU Demand Surge
- Control the Harness, Control the Cost: Routing and Governing AI Coding Agents in the Enterprise
- Matos accuses Guckian of using AI to steal her plan
- Anthropic Quitter Jacob Coxon Worked With PR Firm DEY. Which Has Worked With AI Skeptics Like Eliezer Yudkowsky: Report
- State audit finds gaps in AI oversight at SUNY campuses
- How options traders can play this tech giant's AI-driven breakout
- Curtis Cosponsors Legislation to Protect AI Whistleblowers
- Doomsday Clock expert warns of 'extremely pressing threat' of artificial intelligence
Comments
Please log in to post a comment.