Anthropic, the artificial intelligence firm behind the Claude language model, has moved to strengthen its government relations capacity by appointing Mariano-Florentino Cuéllar as its inaugural chief global affairs officer. The appointment signals the San Francisco-based startup's recognition that managing relationships with regulators and policymakers has become as strategically important as developing its core technology. Cuéllar, who will report to Anthropic President Daniela Amodei, brings substantial diplomatic credentials and bipartisan credibility to a role focused on navigating the fractured political landscape surrounding artificial intelligence regulation.
Cuéllar's background positions him as a bridge-builder between competing ideological camps at a moment when AI governance remains deeply polarised. He served as a special assistant in the Obama administration and most recently led the Carnegie Endowment for International Peace until July, where he co-chaired a task force examining national security implications and nuclear proliferation risks. His appointment reflects Anthropic's understanding that winning policy battles requires more than technical expertise—it demands someone capable of building consensus among ideologically diverse stakeholders who view AI through radically different lenses.
The timing of this hire underscores the mounting pressure on Anthropic from US authorities. The Trump administration has subjected the company to considerable regulatory headwinds this year, including a Pentagon blacklisting of its technology earlier in the year and more recently an export control directive that temporarily prohibited the company from selling its advanced Mythos 5 and Fable 5 models internationally. These restrictions, ostensibly justified on national security grounds, represent the kind of geopolitical friction that Cuéllar will need to manage. Anthropic is currently challenging the Pentagon blacklist in court, indicating the company views these restrictions as potentially overreaching and economically damaging.
Cuéllar's diplomatic approach is already evident in his prior work on artificial intelligence governance. A 2025 study he co-led significantly influenced the development of California's SB 53 legislation, which established requirements for high-revenue AI developers to disclose safety standards and incident reporting while creating whistleblower protections. The law, which Anthropic actively supported, demonstrates Cuéllar's capacity to shape regulatory frameworks in ways that companies in the sector can endorse. This track record suggests he may help steer federal policy in similarly pragmatic directions that balance innovation with genuine oversight concerns.
Cuéllar's hiring also reflects broader industry trends as AI companies confront an increasingly complex regulatory environment globally. Anthropic operates across multiple jurisdictions where policymakers are developing divergent approaches to AI governance. The European Union's AI Act, China's algorithmic regulation requirements, and emerging frameworks across Southeast Asia create a patchwork of rules that companies must navigate. With offices expanding internationally, Anthropic requires someone capable of managing relationships in multiple capitals simultaneously, advocating for policies that don't impose excessive compliance burdens while remaining credible with regulators genuinely concerned about AI risks.
The choice of Cuéllar carries symbolic weight beyond his immediate duties. Conservative analyst Oren Cass, who has worked with Cuéllar on bipartisan policy discussions, characterised him as pragmatic and open-minded toward the Trump administration's positions. This framing is crucial because it signals that Anthropic is not hiring a lobbyist to simply push back against regulation, but rather a figure capable of demonstrating genuine engagement with government perspectives. In polarised Washington, this distinction matters enormously—regulators are more likely to collaborate with firms they perceive as good-faith participants in policymaking rather than entrenched opponents.
Cuéllar's former role on Anthropic's Long-Term Benefit Trust provides additional context for his appointment. This independent governance body exists to ensure the company adheres to its public benefit mission, suggesting that Anthropic structures itself with commitments to responsible development. Cuéllar's transition from this independent oversight position to a company executive role might raise questions about conflict of interest, though his replacement on the trust will presumably maintain the independent character of the body. His move signals confidence that external scrutiny can coexist with internal advocacy for the company's interests.
The appointment arrives as concerns about AI-powered cyber threats intensify. Security vulnerabilities in large language models could be weaponised for sophisticated hacking operations, a risk that national security officials take seriously. This technical backdrop lends legitimacy to government caution around exporting advanced AI capabilities, even as it creates friction between the US government and firms seeking global markets. Cuéllar must convince Trump administration officials that reasonable export policies can coexist with genuine national security safeguards, a delicate balance requiring credibility with both sides.
For Malaysian and Southeast Asian observers, Cuéllar's appointment carries implications for how AI governance will develop in the region. Anthropic's ability to maintain constructive relationships with the US government affects its willingness to invest in emerging markets and the regulatory pressure it experiences globally. Southeast Asian nations developing their own AI frameworks will watch how US regulation develops, and a more collaborative approach between US authorities and leading AI firms could produce models that other countries find more legitimate to adopt. Conversely, if relations deteriorate further, expect more nations to either restrict American AI companies or develop indigenous alternatives.
Cuéllar's statement that AI governance has reached an "inflection point" captures the high stakes involved. The decisions made in the coming months about export controls, safety standards, and transparency requirements will shape which firms thrive globally and which business models become untenable. Anthropic is betting that having a skilled political operator focused exclusively on managing government relationships will help the company thread the needle between innovation and regulation. Whether that proves sufficient depends partly on factors beyond Cuéllar's control—the broader political direction of the Trump administration and whether genuine consensus can emerge around AI governance principles.
