OpenAI is bolstering its roster with high-profile hires as it gears up for its public launch: Noam Shazeer, a legendary figure from Google DeepMind, and Dean Ball, a former AI policy official from the Trump White House. Shazeer, who co-led Google’s Gemini project and founded the AI role-playing startup Character AI, announced his departure on Wednesday. He had been with Google since 2000, leaving only for a three-year period to co-found Character AI. Two years ago, Google brought Shazeer back through a $2.7 billion deal that secured access to the startup’s technology.

This move is the latest in a series of shifts among top AI labs like Google, OpenAI, Anthropic, and Meta. Shazeer is recognized as a foundational figure in modern generative AI, having co-authored the pivotal 2017 paper “Attention Is All You Need,” which introduced the Transformer architecture. Before leaving Google, Shazeer reportedly stirred controversy over political issues. According to The Information, he voiced opinions on internal messaging boards about transgender identity and Israel’s war in Gaza, leading management to delete his posts. Whether these controversies will follow him to OpenAI remains uncertain.

Meanwhile, OpenAI is also strengthening its policy team by bringing Ball on board. Ball had a brief stint in the White House last year, where he helped publish America’s AI Action Plan before stepping down to rejoin the techno-libertarian think tank Foundation for American Innovation as a senior fellow. “I am pleased and honored to announce that, on July 6, I’ll be joining OpenAI as leader of a new team called Strategic Futures,” Ball wrote on X on Thursday. “Our mandate will be to help the company’s leadership shape frontier AI policy.”

Ball will report directly to Chief Strategy Officer Jason Kwon. The “small, high-agency team” will focus on “matters pertaining to: catastrophic risk, recursive self-improvement, labor market impact, and the relationship between the frontier labs, governments (particularly the U.S. Federal Government), and society,” Ball wrote in a blog post. The Strategic Futures team will address both public-facing policy and internal governance, he added. This last point is crucial—Ball noted that “almost by necessity,” AI labs will have to lead on AI governance decisions. “In other words, internal governance will be more central to the future of AI than most people realize,” Ball wrote.

Ball’s decision to join OpenAI—arguably a favorite in the current administration—comes as Anthropic faces renewed conflict with the U.S. government. Late last week, President Donald Trump ordered an export control ban on Anthropic’s latest models, Fable 5 and Mythos 5, forcing the firm to take the models down entirely to avoid noncompliance. For anyone who had “government interference” on their S-1 risk factor bingo card, Ball’s hire illustrates what it looks like when a company secures insider status while a rival faces pressure.