Menu

1 / 2

AI must be held accountable — value statements won’t protect human dignity

Nature·Cindy Skach·about 6 hours ago
#XIE8psSP
Reading 0:00
15s threshold

There is a striking gap in debates about artificial intelligence. Policy bodies and frontier AI laboratories are borrowing the vocabulary of legal constitutions to guide model development and oversight. However, these statements currently don’t include the institutional safeguards that would make such value commitments work.

In September, the UK Parliament’s Joint Committee on Human Rights published its report on AI regulation . It uses ‘human rights’ more than 100 times in as many pages, but doesn’t discuss which rights should be protected or how. AI firm Anthropic in San Francisco, California, has written a ‘ constitution ’ for its model Claude. In it, numerous values are listed as the tool’s priorities, including the preservation of user autonomy. Yet some of these values, such as autonomy and well-being, compete with each other, as Anthropic acknowledges.…

Continue reading — create a free account

Join HashtagPLUS to read full articles, follow hashtags, vote, and join the conversation.