Anthropic CEO Dario Amodei says the most accurate criticism of AI companies is that they have not yet delivered on major promises to improve the world. In posts on X, Amodei argues that public trust will depend on real breakthroughs rather than claims that have not materialized.

The comments are described as part of a broader “crisis of trust,” with Fortune reporting that Amodei highlights concerns that people worry governments or companies may be “cooking up some new way to screw them over.” The Next Web frames the statement alongside Anthropic’s preparations for a potential listing, noting that backers expect a valuation around $2 trillion. While the outlets focus on different implications—one emphasizing trust concerns and the other tying the remark to Anthropic’s market plans—they both present Amodei’s position that delivery against expectations remains a central issue.

Overall, both sources report the same core message: the AI sector has yet to fully meet the public’s expectations, and credibility will hinge on demonstrable results.