The Biden administration secured voluntary commitments from AI companies in 2023 to drive safe, secure, and trustworthy AI development. Those eight commitments were translated into 30 yes-no indicators and 16 signatory companies were scored based on public disclosures through December 31, 2024. With no clear White House guidance, a simple public-evidence test was used: whether any public evidence showed companies acting to fulfill commitments. Approximately half of signatories lacked public evidence of compliance by end-2024. OpenAI, Anthropic, Google, and Microsoft ranked highest while Apple scored lowest at 13 percent, partly reflecting a later joining date.
drive safe, secure, and trustworthy development of AI technology.
This came at a time politically when we expected things might be changing,
It's difficult to discern what exactly constitutes a company satisfying their commitment adequately versus not,
Is there any evidence publicly that the companies are doing something in furtherance of these commitments?
Collection
[
|
...
]