US Senate Majority Leader Thune believes President Trump is open to implementing AI guardrails despite his public defiance on the issue, according to Axios

Episodes of this kind on technology oversight have historically run through the same channel, with an executive order, agency guidance, or a narrower legislative vehicle substituting for comprehensive statute, and predictability on that choice being what matters for the affected sector. The distinction worth drawing is between guardrails framed as export-controls and security screening, which concentrate on the hardware and compute supply chain, and those framed as liability and model-safety rules, which fall on the application layer; the two behave differently for the large-cap concentration in that sector. The actors here matter: a Senate leader conveying openness is usually testing whether a deal is available rather than committing to one. Worth watching is whether the administration's public defiance softens into a framework announcement, and whether any framework reaches the legislative calendar or stays at the executive level. As positioning rather than policy, the signal is soft.