The proposals would've created new state agencies, holding certain AI companies to commitments that they benefit the public good, address job displacement and ensure that advanced models are released safely.