Anthropic faces power struggle with Trump administration over AI regulation

Key Takeaways

Anthropic's Commercial Success Sparks Power Struggle

Anthropic's rapid rise has sparked a power struggle with the Trump administration over AI regulation, national security, and control of advanced models. The clash threatens its IPO, highlights gaps in AI governance, and raises questions about who should control transformative AI technology. Anthropic's CEO, Dario Amodei, has emphasized the need for safeguards on AI, while the Trump administration sees AI as a tool to strengthen America's position in the world.

Anthropic's Mythos AI Breaches NSA Systems

Anthropic's Mythos AI has reportedly breached almost all of the NSA's classified systems in hours, prompting a US-ordered shutdown. The incident highlights the potential risks of advanced AI models and the need for stricter controls. The Trump administration has directed Anthropic to limit access to its Fable 5 and Mythos 5 AI models to US citizens.

AI Inflates Student Grades

A UC Berkeley study found that courses with a high share of writing and coding assignments saw grades jump after ChatGPT launched. The effect shows up mainly in homework, suggesting that AI is replacing student work rather than improving learning. The study warns that grades may be losing their value as a signal if AI-generated output is not properly accounted for.

ClawHub Plugins Expose AI Registry Security Gaps

23 ClawHub plugins have been found to be squatting official scopes, exposing security gaps in the AI registry. The incident highlights the need for better security measures in AI registries.

Jupiter Supercomputer Runs on NVIDIA Chips

Europe's first exascale supercomputer, JUPITER, is powered by NVIDIA Grace Hopper Superchips and NVIDIA Quantum-X800 InfiniBand networking.

Young People Sound Alarm on AI

Young people are pushing back on AI hype, demanding that technology be built thoughtfully. They are concerned about the ethics, equity, and environmental impact of AI and want to ensure that technology is developed in a way that benefits everyone.

IBM Partners with OpenAI on Enterprise Security AI

IBM has partnered with OpenAI to bring enterprise-grade security to OpenAI's AI technology. The partnership aims to develop new security features for AI systems.

World Cup Scams Get Harder to Spot

Scammers are using AI to create more convincing fake websites and impersonate legitimate companies to sell fake tickets. Fans need to be aware of the risks and take steps to protect themselves from these scams.

AI's Energy Future Depends on Space Power

The growing demand for AI energy has sparked interest in harnessing continuous solar energy from space. Space-based solar energy could provide a critical new domain for energy access and meet growing demand.

AI in Migration is More Common Than You Think

AI-related technologies have been used in migration systems for decades, even if not always described as AI. Policymakers need to engage with AI directly and consider its implications for migration governance.

Intel and AMD Unveil New x86 Standard for AI

Intel and AMD have unveiled a new x86 standard, ACE, to make CPUs better at running AI models. The standard aims to improve the efficiency of AI workloads on CPUs and reduce the need for data movement between CPUs and GPUs.

Sources

• Anthropic’s astonishing commercial success makes it a target
• Anthropic's Mythos AI broke into almost all NSA classified systems in hours
• AI is inflating student grades, and the effect points to outsourced work, not better learning
• 23 ClawHub plugins squatting official scopes expose AI registry security gaps
• NVIDIA Blog
• Young People Are Sounding the Alarm on AI. We Should Listen to Them
• IBM partners with OpenAI on enterprise security AI
• World Cup Scams Are Getting Harder to Spot
• Why AI's energy future depends on power from space
• The use of AI in migration is more common than you may think
• Intel and AMD unveil new x86 standard to make CPUs better at running AI models