# Dario Amodei's Oversight Plan for AI, and How Altman Differs _Dario Amodei published an essay calling for government power to block dangerous AI deployments. Sam Altman negotiated changes to GPT-5.6 with officials instead. Here's what separates the two approaches in 2026._ **Published:** 2026-07-27 **Source:** https://www.startuphub.ai/ai-news/ai-figures/2026/figure-dario-amodei-strategic-position-vs-peer-2026-07-27 --- **When Dario Amodei published a 5,000-word essay in June 2026 calling on governments to gain statutory power to block AI model deployments, he was making an argument that no sitting AI CEO had put so directly: third-party auditors, not companies, should hold the final veto. Six weeks later, Bloomberg reported that OpenAI's Sam Altman made "many changes" to GPT-5.6 through a "collaborative back and forth" with the Trump administration before its release. The two most closely watched executives in frontier AI have now staked out structurally different positions on the same question, and both are raising money at record valuations while doing so.** Quick context [Dario Amodei](/people/dario-amodei) co-founded [Anthropic](/startups/anthropic) in 2021 after leaving OpenAI, where he served as VP of Research. Anthropic disclosed a $47 billion annualized revenue run rate in May 2026 per [Bloomberg](https://www.bloomberg.com/news/articles/2026-05-28/anthropic-raises-at-965-billion-valuation-eclipsing-openai), the same month it closed a $65 billion round at a $965 billion valuation. The governance debate covered here is drawn from Amodei's June 2026 public essay and a contemporaneous Bloomberg account of Altman's regulatory engagement. ## The Government Veto Amodei Is Proposing Amodei's June 2026 essay, "Policy on the AI Exponential," published on his personal site and covered in an exclusive interview with [ABC News](https://abcnews.com/Business/exclusive-anthropic-ceo-calls-stronger-regulation-ai/story?id=133753620), argued that AI safety measures are structurally out of step with the pace of capability development. His central proposal: governments should have the power to "block or deter deployment of the model if it is determined, in light of third-party assessment, to present unacceptable risks." The mechanism he described requires mandatory independent audits of frontier AI systems before public release, with the blocking authority triggered by audit findings rather than by any one agency's discretion. The proposal was accompanied by a $200 million commitment to fund research on AI's societal impact, per [Axios](https://www.axios.com/2026/06/10/anthropic-ceo-government-block-dangerous-ai). Amodei's framing was explicit about urgency: he described the window between now and 2030 as a period in which AI capabilities could reach what he called a "country of geniuses in a datacenter" level of output, and argued that regulatory institutions need the teeth to act before that inflection, not after. A June 17 [CNBC report](https://www.cnbc.com/2026/06/17/anthropic-ai-regulation-trump.html) noted that Washington's actual response went beyond anything Anthropic formally requested, suggesting the essay landed at a moment when policymakers were already looking for a framework. Critics have characterized the blocking-power proposal as a mechanism that would favor well-capitalized incumbents over smaller entrants, since large labs are better positioned to fund the audits and compliance infrastructure a mandatory review regime implies. Amodei has not publicly addressed the incumbency critique. ## Altman's Approach: Negotiate Before You Release [Sam Altman](/people/sam-altman) has called global AI regulation "urgently" needed, but his operating practice in 2026 has been consultative rather than statutory. [Bloomberg reported on July 9](https://www.bloomberg.com/news/articles/2026-07-09/altman-says-openai-made-many-changes-during-talks-with-us) that Altman described a "collaborative back and forth" with Commerce Secretary Howard Lutnick and Treasury Secretary Scott Bessent in the weeks before OpenAI released GPT-5.6. The company made "many changes" to the model during those conversations, Altman said. On July 21, [Bloomberg reported](https://www.bloomberg.com/news/articles/2026-07-21/openai-s-altman-to-brief-us-officials-on-next-wave-of-ai-models) that Altman is scheduled to brief US officials again on OpenAI's next generation of models, suggesting the bilateral pre-release review has become standard practice between the company and Washington. The structural difference from Amodei's proposal is significant. Under the Altman model, authority rests with the company and the executive branch, filtered through an ongoing relationship. Under Amodei's proposal, authority sits with a third-party assessor backed by statute, and the company cannot proceed without clearance from that body. Both positions claim to target the same outcome: preventing genuinely dangerous AI from reaching the public. They differ on who holds the blocking power and on what basis it can be exercised. The two diverge on AGI framing as well. Amodei has projected that AI systems reaching "country of geniuses" capability levels could arrive by 2026 or 2027. Altman, meanwhile, has said AGI is "not a super useful term" even while OpenAI has raised capital explicitly on its AGI mission, per reporting cited in our [July 26 profile of Altman](/ai-news/ai-figures/2026/figure-sam-altman-venture-portfolio-breakdown-2026-07-26). The framing gap matters for governance: if AGI is imminent, Amodei's blocking authority is urgent; if it is indefinite, Altman's consultative model has more runway. ## Safety Positioning as Capital Thesis Anthropic's policy stance is not operating independently of its fundraising. The $65 billion round at a $965 billion valuation, announced May 28, 2026 and reported by [Bloomberg](https://www.bloomberg.com/news/articles/2026-05-28/anthropic-raises-at-965-billion-valuation-eclipsing-openai), came with the same messaging around responsible deployment that defines Amodei's public essays. Investors pricing Anthropic above OpenAI's $852 billion last known valuation are incorporating the safety narrative into their thesis, whether explicitly or as brand value in enterprise sales. The July 2026 Ode with Anthropic joint venture makes that connection concrete. Blackstone and Hellman and Friedman committed $1.5 billion to an enterprise AI services firm built explicitly on Anthropic's models, per [BusinessWire](https://www.businesswire.com/news/home/20260715100000/en/Anthropic-Blackstone-and-Hellman-Friedman-Introduce-Ode-with-Anthropic-an-Enterprise-AI-Services-Firm). Enterprise buyers selecting Claude over GPT are partly purchasing the safety positioning; private equity firms allocating $1.5 billion alongside that choice are pricing it into a return model. The Ode JV is one of the clearest signals that large institutional capital views Anthropic's regulatory posture as a commercial differentiator, not a constraint on growth. StartupHub.ai data shows that among the six frontier AI labs tracked in our startup database, Anthropic and xAI share a composite score of 76, below OpenAI's 84. The gap reflects OpenAI's larger commercial footprint and product distribution rather than a technical or safety ranking. Across all six, only Anthropic has proposed statutory government blocking authority as a formal deployment condition, a policy position now backed by a $965 billion valuation story. ## What It Means Amodei and Altman are not arguing about whether AI should be governed. Both have called for oversight and both are actively engaged with regulators. The disagreement is architectural: one wants the veto power sitting inside government, triggered by independent auditors before a model can ship; the other operates through pre-release negotiation with executive branch officials, with the company retaining the final decision. The consultative model is what exists in practice in Washington today. Amodei's proposal is a bet that voluntary engagement is insufficient for the capability levels coming in the next two to three years, and that whoever builds the case for binding institutional oversight earliest will shape the regulatory frame that eventually governs the entire industry. ## Sources - ![](https://www.google.com/s2/favicons?domain=bloomberg.com&sz=32) [Bloomberg: Anthropic Raises at $965 Billion Valuation Eclipsing OpenAI](https://www.bloomberg.com/news/articles/2026-05-28/anthropic-raises-at-965-billion-valuation-eclipsing-openai) - ![](https://www.google.com/s2/favicons?domain=bloomberg.com&sz=32) [Bloomberg: Altman Says OpenAI Adjusted AI Models After Discussions With US Officials](https://www.bloomberg.com/news/articles/2026-07-09/altman-says-openai-made-many-changes-during-talks-with-us) - ![](https://www.google.com/s2/favicons?domain=bloomberg.com&sz=32) [Bloomberg: OpenAI's Altman to Brief US Officials on Next Wave of AI Models](https://www.bloomberg.com/news/articles/2026-07-21/openai-s-altman-to-brief-us-officials-on-next-wave-of-ai-models) - ![](https://www.google.com/s2/favicons?domain=abcnews.go.com&sz=32) [ABC News: Anthropic CEO Calls for Stronger Regulation of AI (exclusive)](https://abcnews.com/Business/exclusive-anthropic-ceo-calls-stronger-regulation-ai/story?id=133753620) - ![](https://www.google.com/s2/favicons?domain=axios.com&sz=32) [Axios: Anthropic CEO says government should block dangerous AI](https://www.axios.com/2026/06/10/anthropic-ceo-government-block-dangerous-ai) - ![](https://www.google.com/s2/favicons?domain=cnbc.com&sz=32) [CNBC: Anthropic asked for regulation. Washington went much further.](https://www.cnbc.com/2026/06/17/anthropic-ai-regulation-trump.html) - ![](https://www.google.com/s2/favicons?domain=businesswire.com&sz=32) [BusinessWire: Anthropic, Blackstone and Hellman & Friedman Introduce Ode with Anthropic](https://www.businesswire.com/news/home/20260715100000/en/Anthropic-Blackstone-and-Hellman-Friedman-Introduce-Ode-with-Anthropic-an-Enterprise-AI-Services-Firm) - ![](https://www.google.com/s2/favicons?domain=darioamodei.com&sz=32) [Dario Amodei: Policy on the AI Exponential](https://darioamodei.com/post/policy-on-the-ai-exponential) Editorial standards: every claim is sourced. Tips: editor@startuphub.ai --- Original analysis from [startuphub.ai](https://www.startuphub.ai), the #1 AI startup directory.