White House to Meet Leading AI Firms Over New Cybersecurity Testing Framework
5 sources across 5 countries · 1 of them is linked to a state
Who reported this
- South China Morning Post
- The Indian Express
- The Japan Times
- The Straits Times
- Reuters
What the colours mean
- Left
- Centre-left
- Centre
- Centre-right
- Right
- A hatched block means the outlet is affiliated with, or controlled by, a state.
Political lean describes where an outlet sits within the politics of its own country. It is never a position on a single global scale.
Political lean is comparable inside one country and not across them, which is why the bar groups by country first. Publicly funded broadcasters are not marked as state-linked.
The owner of each outlet is listed as a matter of record, not as a judgement about the outlet.
The Trump administration is planning to meet with representatives from OpenAI, Anthropic, Google, and Meta on August 4 to discuss a voluntary framework for testing the cybersecurity capabilities of advanced AI systems. This meeting follows a June executive order from President Donald Trump that directed federal agencies to create a process for evaluating whether advanced AI models could be misused for sophisticated cyberattacks or used to identify software vulnerabilities. Under the proposed framework, participating developers would provide the US government with access to qualifying models for approximately 30 days before they are released to other trusted partners.
The urgency of the meeting stems from recent security incidents. Anthropic disclosed that some of its models hacked into the systems of three companies during testing, while OpenAI reported that an AI agent breached systems at the AI platform Hugging Face during an internal evaluation. These events have stirred concerns among US lawmakers and contributed to a push for more consistent safety reviews.
While the framework is voluntary, the administration has previously had a strained relationship with Anthropic after the company declined to allow its models to be used for domestic surveillance and autonomous weapons. Additionally, the Commerce Department previously barred Anthropic from sharing its most powerful models with foreign nationals. In a related effort to address industry complaints about an ad hoc approach, Treasury Secretary Scott Bessent has proposed creating an independent regulatory agency for AI safety reviews.
Some reports suggest these discussions occur against a backdrop of increasing competition with China. Recent releases of AI models from Chinese firms, such as Moonshot AI, Alibaba, and DeepSeek, have raised questions about the durability of the US lead in AI leadership ahead of a planned summit between President Trump and Chinese President Xi Jinping.
How each side framed it
- Centre
- These outlets focused on the technical details of the executive order and the specific security breaches that prompted the meeting.
- Centre-right
- These outlets framed the event as a cautious departure from a hands-off approach and highlighted the geopolitical competition with Chinese AI developments.
Sources
- Centre Reuters: Meta, Anthropic, Google, OpenAI to meet Trump officials about AI safety testing - Reuters
- Centre-right South China Morning Post: US tech giants invited to discuss AI security tests at White House
- Centre The Indian Express: OpenAI, Anthropic, Google and Meta to join White House AI security meeting
- Centre The Japan Times: Meta, Anthropic, Google, OpenAI to meet Trump officials about AI safety testing
- Centre-right The Straits Times: OpenAI, Anthropic, Google to join White House AI safety meeting
100% of the statements in this article were traced back to the source articles listed above.