Which AI companies ranked highest and lowest in the new safety assessment?

Question

Accepted Answer

Anthropic and Mistral Ranked Highest and Lowest, Respectively, in New AI Safety Assessment

Anthropic secured the top position in a recent global AI safety ranking, while Mistral finished last among the assessed companies. The evaluation, conducted by the Future of Life Institute, involved nine leading AI firms and aimed to gauge their preparedness for "existential" risks posed by advanced artificial intelligence. No company achieved an "A" grade in any of the six assessed categories, with Anthropic receiving the highest overall score of "C+".
Anthropic Leads Overall Safety Ranking

US-based AI lab Anthropic emerged as the highest-scoring company in the new global AI safety assessment, earning an overall grade of "C+". The institute's researchers evaluated nine major AI companies across six categories: risk assessment, current harms, safety frameworks, existential safety, governance and accountability, and information sharing. Despite its top ranking, Anthropic was noted for "questionable military engagements," with reports indicating its technology was used in military operations involving Venezuela and Iran.
Mistral Disputes Assessment Framework

France's Mistral, participating in the rankings for the first time, was placed last. The company contested the assessment, asserting that the report's framework was not suitable for its open-model development approach. Unlike closed-model developers, Mistral focuses on open models that users can download and modify, a strategy it believes was not adequately reflected in the evaluation criteria. The Future of Life Institute noted that Mistral did not respond to their survey for data.
Meta and xAI Show Varied Performance

Meta demonstrated an improvement in the rankings, climbing two places to secure the fourth position. In contrast, xAI experienced a significant drop, falling three spots to seventh place. These shifts highlight the dynamic nature of AI safety evaluations and the varied progress among prominent technology companies. Both Alibaba and DeepSeek also featured in the lower half of the rankings, with Alibaba Cloud at sixth and DeepSeek at fifth.
Industry-Wide Shortcomings Highlighted

The report concluded that all nine companies assessed are inadequately prepared to address the "existential" threats associated with artificial general intelligence (AGI). While acknowledging the existence of "constructive attempts," the researchers deemed industry-wide efforts as "entirely inadequate." Concerns were also raised about the potential misuse of advanced AI for cyberattacks or other harmful activities. Several companies that had previously committed to avoiding military applications have reportedly softened these stances.

Reading history

Hide →

This history is cleared when the page is closed.

Which AI companies ranked highest and lowest in the new safety assessment?

Loading

Was this article helpful?

Helpful

Not helpful

Report an issue

2

0

0

Share with friends