Trump’s AI testing plan is limited and vague
The Trump administration's AI testing framework excludes open models and cannot restrict them post-release, according to Axios reporting on a voluntary guideline developed after an executive order.
- The White House's AI testing framework excludes open models from cybersecurity assessments.
- The framework cannot restrict open models after their release, limiting its regulatory scope.
- The voluntary guidelines were developed after an executive order requiring AI companies to share frontier models with the government.
- Critics highlight potential oversight gaps due to the exclusion of open models.
The White House's AI testing framework, created following a June executive order, explicitly excludes open models from its voluntary cybersecurity assessments. According to Axios, the guidelines state that open models, those available for public download and inspection, are not subject to the framework's evaluation process. Furthermore, the framework cannot be used to impose restrictions on open models after they have been released, limiting its ability to address potential risks associated with widely accessible AI systems.
The framework was developed in response to President Trump's directive for AI companies to share their frontier models with the federal government prior to deployment. While the move aims to enhance national security by assessing risks in closed, proprietary models, the exclusion of open models raises concerns about oversight gaps. Critics argue that this limitation could leave significant vulnerabilities unaddressed, particularly as open models become increasingly powerful and widely used in critical applications.
Open model developers may face less scrutiny, but the framework's limitations could impact trust in AI systems.
Companies relying on open models may find the framework's exclusion beneficial, but broader regulatory uncertainty remains.
The framework's narrow scope could influence investment decisions in AI sectors, particularly those involving open models.
The policy raises questions about the balance between innovation and security in AI governance.
- open models
- AI models whose code, weights, and architecture are publicly available for inspection and modification.
- frontier models
- Cutting-edge AI systems that represent the current state of the art in capabilities.
AI ResearchWeatherNext: AI model achieves breakthrough in forecasting cyclones
DeepMind introduced WeatherNext, an AI system that markedly improves cyclone track and intensity predictions, extending forecast lead times by several days.
US Senate Commerce approves KOSA, children's AI safety bills - IAPP
The US Senate Commerce Committee has approved two bills focused on AI safety for children. The bills aim to regulate AI systems and protect children's data.
Artificial intelligence enters Italy’s national security agenda - Decode39
Italy has added artificial intelligence to its national security agenda, marking a significant development in the country's approach to AI. This move is expected to have implications for the nation's defense and security strategies.
Powering the ballot: Why AI’s energy footprint is the ultimate midterm election issue - Route Fifty
AI’s growing energy demands are becoming a key issue in the US midterm elections, raising questions about sustainability and infrastructure.
Accelerating Biomedical Innovation with AI through Collaborative Iteration - Wyss Institute at Harvard
The Wyss Institute at Harvard is leveraging AI to accelerate biomedical innovation through collaborative iteration. Researchers are using AI to analyze and improve medical devices and treatments.
DeepSeek invests $20.8 million in Unitree's Shanghai IPO - Reuters
DeepSeek has committed $20.8 million to Unitree's upcoming Shanghai IPO, signaling strong investor confidence in the robotics firm.