Cruz calls out nonprofit that received federal grants for ‘hiding’ work to shape AI policy

The bills irked free-market and consumer groups alike. They generally required AI developers to share information about their training data and processes with AI deployers who would then be required to complete detailed assessments of their systems for algorithmic bias and discrimination risk, prompting concerns they would slow innovation and adoption of the technology by new entrants.

The bills also all included loopholes waiving the requirements if there was “sufficient human review” — consumer groups argued this would sanction one human in the loop simply rubberstamping an AI’s outputs — while degrading the private right of action.

Although the state working group’s efforts were also actively endorsed by the human resources giant Workday, which is being sued for alleged discrimination in conjunction with a tool it provides for hiring decisions, FPF became the focus of groups like the R Street Institute and the Mercatus Center…

Following a Jan. 24 social media confrontation in which the R Street Institute’s Adam Thierer directly asked Polonetsky about FPF’s role promoting the legislation, the FPF CEO defended the group’s work describing its role as a “neutral facilitator.” But in a Feb. 25 post to the site, he said FPF was ending its engagement with the working group, citing a desire to continue being perceived as politically neutral.

The state policymaking effort has since lost some momentum, with Virginia Gov. Glenn Youngkin (R) vetoing that state’s proposal and Texas state Rep. Giovanni Capriglione (R) on March 14 quietly introducing a revised version of his legislation which massively curtails its original provisions.

“New developments in Virginia and Texas signal that the debate over artificial intelligence (AI) policy could be turning in a more positive, pro-innovation direction in the states,” Thierer wrote in a March 25 piece noting the changes.