AI CEOs Skip Trump EO, Policy in Limbo

Trump EO stalls after AI CEOs boycott

The White House announced an executive order on AI safety testing, then cancelled the signing ceremony. The cancellation came after CEOs of leading AI firms refused to attend. The administration framed the order as a safeguard, calling it a needed check on risky models. The CEOs said the order would act as an “innovation blocker” and walked out. No official list of the firms was released, but the media identified the biggest players in the field. The abrupt change left the policy in limbo and raised questions about the administration’s leverage over the industry. The episode underscores a growing friction between regulators and private AI labs. The order promised mandatory safety audits before deployment, but the industry argues that the process is undefined and could delay product launches. Without the CEOs’ buy‑in, the White House now faces a stalled rollout and a political optics problem.

Mozilla rebrands amid AI pressure

The Mozilla Foundation unveiled a new visual identity in August, and its for‑profit arm, Mozilla Corporation, refreshed its logo in December. The redesign coincided with Nabiha Syed’s appointment as executive director in May. Syed told The Register that the foundation remains committed to an open, inclusive internet. She warned that “the onslaught of AI slop and surveillance capitalism” forces a return to first principles. The rebrand is meant to signal resilience after layoffs and to position Mozilla as a watchdog in the AI era. She emphasized that AI is the next frontier of digital life, extending the historic shift from browsers to social feeds to generative models. “They’re all part of a digital experience,” she said. Mozilla plans to lean on its existing projects, like Common Voice, to counterbalance corporate data collection. Common Voice aggregates speech samples in dozens of languages, including low‑resource tongues such as Flemish, Catalan, and Sindhi. Companies like Meta and Nvidia already use the dataset, and activists have built niche chatbots on it. Syed argues that open data can keep AI development from consolidating power in a few hands.

Google curtails AI overviews on Trump health queries

Google’s AI Mode now returns a list of ten links instead of a generated summary for the query “does trump show signs of dementia.” The same query for former President Joe Biden yields an AI Overview that begins with a disclaimer about the lack of definitive evidence. The inconsistency extends to other presidents. A search for Barack Obama’s health produces a concise AI Overview stating no public evidence of dementia. Google’s spokesperson Davis Thompson replied that “AI Overviews and AI Mode won’t show a response to every query” and cited internal guidelines for omission. The pattern suggests a selective filtering of politically sensitive content. The move follows a recent $24.5 million settlement over a lawsuit that claimed Trump’s YouTube account was improperly banned. Critics argue that the AI omission may be a pre‑emptive shield against potential legal or reputational fallout.

Industry context: corporate gatekeeping and policy tug‑of‑war

These three developments illustrate a broader struggle over who controls AI narratives. The Trump administration’s attempt to impose safety audits met immediate pushback from the very firms it sought to regulate. Mozilla’s rebranding signals a nonprofit’s attempt to reclaim the public‑interest space that large AI companies dominate. Google’s selective AI responses add another layer. By withholding AI‑generated answers on a high‑profile health question, the company effectively cedes the narrative to traditional search results, which are more easily curated. This tactic mirrors earlier content moderation debates, but now applies to AI‑generated prose. The common thread is a reluctance to let unfiltered AI output shape public perception without oversight. Companies cite accuracy concerns; regulators cite safety. The tension creates a de‑facto gatekeeping role for the industry, even as policymakers push for more formal controls.

What to watch

Watch the White House’s next move on the AI safety order. A revised draft or a new enforcement timeline could signal whether the administration will compromise or double down. Track Mozilla’s upcoming policy proposals, especially any that target AI data collection or model transparency. Finally, monitor Google’s AI Mode documentation for changes to the exclusion criteria, and watch for any formal complaints or legal actions that could force the company to standardize its AI overview policies across all political figures.