09/17/2026
๐๐-๐บ๐ถ๐ป๐ด ๐๐ผ๐ ๐ผ๐ป ๐๐ ๐๐ฐ๐ฐ๐ผ๐๐ป๐๐ฎ๐ฏ๐ถ๐น๐ถ๐๐
In an excellent article from FT financial commentator and next James Bond Robert Armstrong, he writes โTrust is built on accountability, but the AI companies โ from hyperscalers like Google to model builders like OpenAI and Anthropic โ have been unable and unwilling to make themselves accountable.โ
It's perhaps more accurate to say that they have been unwilling to be fully transparent.
I draw the distinction because ๐ฎ๐๐๐ต๐ฒ๐ป๐๐ถ๐ฐ ๐ฎ๐ฐ๐ฐ๐ผ๐๐ป๐๐ฎ๐ฏ๐ถ๐น๐ถ๐๐ ๐ฟ๐ฒ๐พ๐๐ถ๐ฟ๐ฒ๐ ๐๐ผ๐บ๐ฒ ๐ฒ๐
๐๐ฒ๐ฟ๐ป๐ฎ๐น ๐ฒ๐ป๐ณ๐ผ๐ฟ๐ฐ๐ฒ๐บ๐ฒ๐ป๐ ๐บ๐ฒ๐ฐ๐ต๐ฎ๐ป๐ถ๐๐บ.
By way of example, the National Transportation Safety Board (NTSB) has the legal authority to investigate civilian aviation accidents. It is legally responsible for civil aviation accident oversight.
While the NTSB still maintains final authority and oversight, it frequently delegates initial fact-gathering or field investigations to the Federal Aviation Administration (FAA).
But the FAA also has the legal regulatory power to ground aircraft, issue emergency airworthiness directives, or suspend pilot certificates.
๐ช๐ต๐ผ ๐ต๐ฎ๐ ๐๐ต๐ฒ ๐ฝ๐ผ๐๐ฒ๐ฟ ๐๐ผ ๐๐๐๐ฝ๐ฒ๐ป๐ฑ ๐๐ ๐บ๐ผ๐ฑ๐ฒ๐น๐ ๐๐ต๐ฎ๐ ๐ต๐ฎ๐๐ฒ ๐๐ฎ๐ธ๐ฒ๐ป ๐ฑ๐ฎ๐ป๐ด๐ฒ๐ฟ๐ผ๐๐ ๐ฎ๐ฐ๐๐ถ๐ผ๐ป๐?
Nobody.
And until that gap in oversight is addressed, any talk of accountability in the AI industry is just that - talk...
Here's the link to Rob's article: https://www.ft.com/content/f5325951-7049-4d3a-97e4-cbe34f9058d8?accessToken=zwAAAaCvY8BokdP1MllRcElNOtOX5MvjT5BY2A.MEUCIFS_D-gdSth-nCGWCaa5CEM-W_2r36UUw0tO2tXjPdYIAiEA0VO-9CsQIEH4dW5esb1ZqsokM3c8MLauGfG3N8JN7qk&segmentId=e95a9ae7-622c-6235-5f87-51e412b47e97&shareId=b63376a8-1751-4e04-b514-3b62ec3174fc&shareType=enterprise&syn-25a6b1a6=1
OpenAI Anthropic Claude Community Mistral Meta