TLDRPresident Trump has issued an executive order establishing voluntary early federal access to cutting-edge AI systems for up to 30 days prior to public launch.Security concerns surrounding Anthropic’s Mythos model, which demonstrates unprecedented cybersecurity exploitation capabilities, catalyzed the order.Initial drafts proposed a 90-day evaluation period; industry opposition advocated for only 14 days, resulting in the 30-day compromise.The directive establishes a collaborative “AI cybersecurity clearinghouse” uniting government entities with private sector partners.Participation remains entirely voluntary, with the order explicitly rejecting any compulsory licensing or permit requirements for AI developers.The executive order received Trump’s signature on June 3, 2026, following multiple postponements. Below is a comprehensive breakdown of its provisions and implications.President Trump signs executive order to review advanced AI models.The administration will ask leading AI developers to voluntarily submit their most powerful AI models for cybersecurity testing before release, and agencies would get up to 30 days to test them.The policy idea… pic.twitter.com/yt6DkAFPRs— Rohan Paul (@rohanpaul_ai) June 3, 2026Anthropic’s Mythos Model: The Catalyst Behind Federal ActionAnthropic’s unreleased Mythos model remains unavailable to the general public. According to company statements, this system demonstrates cybersecurity vulnerability exploitation capabilities at velocities that have triggered alarm among both federal authorities and financial sector observers.The primary apprehension centers on advanced AI systems potentially accelerating cyberattack effectiveness by enabling malicious actors to identify and leverage security weaknesses in essential infrastructure—including financial institutions, healthcare facilities, and governmental networks—outpacing defensive countermeasures.Anthropic has announced plans to distribute Mythos across its customer base within weeks. This pre-order announcement heightened the urgency surrounding the executive action.Core Provisions of the Executive OrderThe directive permits AI companies to voluntarily provide federal agencies with new model access up to 30 days before external distribution to partners or public consumers.This 30-day period represents a negotiated middle ground. A May preliminary version requested 90 days of advance government access. Technology firms mounted significant resistance, contending that 14 days better aligned with the industry’s rapid development pace.The order explicitly clarifies that no mandatory licensing or authorization framework is being imposed. Company involvement remains entirely discretionary.Additionally, it mandates the Treasury Department, National Security Agency, and Cybersecurity and Infrastructure Security Agency to establish a unified “AI cybersecurity clearinghouse” dedicated to identifying and addressing software vulnerabilities through public-private collaboration.Industry Leaders RespondOpenAI’s Sam Altman characterized the order as achieving “the right balance.” Google’s Kent Walker described it as “an important step forward.” Anthropic, despite previous tensions with the Trump administration, endorsed the directive, stating it represents “an important step in strengthening America’s leadership in AI.”Microsoft President Brad Smith praised the initiative, noting it balances innovation advancement with public security protection.The signing follows a turbulent period. The White House came close to announcing the order on May 20, but Trump abruptly delayed it, expressing dissatisfaction with “certain aspects” and concerns about potentially hindering AI advancement.David Sacks, Trump’s AI and cryptocurrency advisor, reportedly contacted the president directly to caution that the initial draft risked stifling innovation and weakening America’s competitive position against China in artificial intelligence.Skeptics Voice ConcernsUniversal approval remains elusive. Anthony Aguirre, CEO of the Future of Life Institute, acknowledged the order as “an important step in the right direction” while cautioning that voluntary mechanisms prove insufficient. He contends federal authorities require authority to prohibit AI systems presenting unacceptable national security threats.The European Union’s AI Act, implemented in 2024, establishes legally binding standards for high-risk artificial intelligence, encompassing mandatory safety evaluations and incident disclosure requirements. The American regulatory approach remains considerably more permissive by contrast.Trump eliminated a Biden-era AI oversight directive on his first day returning to office. Biden’s 2023 order similarly depended primarily on voluntary industry commitments.Anthropic is anticipated to move forward with Mythos model deployment to its customer base in upcoming weeks, now operating under this newly established voluntary evaluation framework.The post Trump’s New AI Executive Order Grants Federal Government 30-Day Preview Period for Advanced Models appeared first on Blockonomi.