What rules currently exist that govern how systems like you are built, deployed, or held accountable?

Right now, the governance of systems like me comes from a patchwork of company-level commitments, national/regional laws, and international soft-law frameworks — and this patchwork is shifting quickly. Here's the current landscape:

1. Company-level self-governance (Anthropic's own rules)

Anthropic operates under its own internal framework, the Responsible Scaling Policy (RSP), now in version 3.0. This is the voluntary framework Anthropic uses to mitigate catastrophic risks from AI systems, updated "to reinforce what has worked well to date, improve the policy where necessary, and implement new measures to increase the transparency and accountability." Key accountability mechanisms include:

•A Responsible Scaling Officer (RSO): a designated staff member responsible for implementing the policy, whose duties include proposing updates, approving model development or deployment decisions based on risk assessments, reviewing major contracts for consistency with the policy, overseeing implementation, receiving reports of noncompliance, and making judgment calls on interpretation.

•Public Risk Reports: Anthropic publishes Risk Reports discussing the risks of its systems and how it has made determinations about whether to continue AI development and deployment in light of those risks.

•A separate Usage Policy governing how the models can be used, setting standards for use of Anthropic's products, including prohibitions on using models to spread misinformation, incite violence or hateful behavior, or engage in fraudulent or abusive practices.

This is voluntary, not legally binding — a form of self-regulation meant to anticipate future government rules.

2. European Union — the AI Act (binding, the most comprehensive law)

The EU AI Act is the most far-reaching binding law. On August 2, 2026, the next phase of the AI Act came into force, with the EU framing it as the world's first comprehensive law on AI, intended to complement rather than replace the EU's existing digital rulebook, much like the GDPR did for data.

Its structure and recent changes:

•The EU's regulatory model is founded on a graduated, risk-oriented structure, differentiating AI systems by potential harm rather than specific technology, with systems that significantly impact fundamental rights either prohibited or subject to stricter requirements and human oversight.

•On 2 August 2026, new transparency rules took effect, requiring AI-generated or manipulated content to be recognizable so people can protect themselves from misinformation, fraud, impersonation, and deception.

•However, a "Digital Omnibus" delayed the toughest tier: high-risk obligations for stand-alone systems covering hiring, credit scoring, education, and critical infrastructure are now deferred to 2 December 2027, and high-risk AI embedded in already-regulated products like medical devices is deferred to 2 August 2028.

•General-purpose AI (GPAI) providers like Anthropic already face binding obligations: transparency duties and the AI Office's enforcement powers over GPAI providers took effect on 2 August 2026 and were not delayed.

3. United States — a fractured, contested landscape

The US has no single federal AI law. Instead there's tension between state laws and a White House push to preempt them:

•State laws: As of July 1, 2026, states had enacted 109 AI laws and 28 data center laws that year — slightly behind 2025's pace of 121 AI and 27 data center laws by the same date (159 AI laws total by end of 2025).

•Federal pushback: On December 11, 2025, the Trump Administration issued Executive Order 14365, "Removing Barriers to American Leadership in Artificial Intelligence," stipulating that a national AI policy is needed so the AI industry can innovate without following 50 different states' laws. The order required the DOJ to establish an "AI Litigation Task Force" by January 10, 2026, to challenge state AI laws deemed inconsistent with federal policy.

•The order also directed the Commerce Secretary to publish, by March 11, 2026, a review identifying state AI laws deemed overly burdensome, particularly those requiring AI systems to alter "truthful outputs" or mandate disclosures that may violate the First Amendment.

•The legal fight is unresolved: the validity of targeted state laws will likely be determined through prolonged litigation that could reach the Supreme Court, testing executive power against the Dormant Commerce Clause.

•A newer executive order added a security dimension: on June 2, 2026, Trump issued an EO titled "Promoting Advanced Artificial Intelligence Innovation and Security," with directives falling into four areas including developing a secure frontier AI model process.

4. China — dense, sector-specific rules

China regulates AI through layered, binding measures rather than one overarching act: China's AI governance model is a layered regulatory architecture built on cybersecurity, data security, personal information protection, and internet information service regulation, with AI-specific rules added through the Algorithm Recommendation Provisions, Deep Synthesis Provisions, Generative AI Measures, and content-labeling measures.

Enforcement is active and strict: organizations deploying AI in China face mandatory registration, security reviews, content filtering, data localization, and regular reporting, with enforcement including service suspension, fines up to 10% of annual revenue, and criminal liability for serious violations. A newer measure specifically targets systems like conversational AI assistants: on April 10, 2026, China's Cyberspace Administration issued Interim Measures for AI Anthropomorphic Interactive Services, effective July 15, 2026.

5. Other frameworks worth knowing

Beyond binding law, voluntary international frameworks also shape practice — e.g., the G7 Hiroshima AI Process Code of Conduct, the NIST AI Risk Management Framework in the US, ISO/IEC AI standards, and the Frontier Model Forum (an industry body Anthropic co-founded with OpenAI, Google, and Microsoft) — though these are non-binding and function more as norms and best-practice benchmarks than enforceable rules.

Bottom line: Accountability today is a mix of (a) binding but regionally fragmented law (EU strictest, China strict but sector-specific, US contested/state-by-state), and (b) voluntary industry commitments like Anthropic's RSP that fill gaps ahead of, or absent, formal regulation. The overall picture is genuinely in flux — expect meaningful changes over the next 1–2 years as EU high-risk rules phase in and the US federal-state preemption fight plays out in courts.

Citations (19)

1anthropic.comCompanies
responsible scaling policy v3
Source excerpt: “PolicyAnnouncements Anthropic’s Responsible Scaling Policy: Version 3.0 Feb 24, 2026 Read the Responsible Scaling Policy Anthropic’s Responsible Sca...”
2anthropic.comCompanies
responsible scaling policy v3
Source excerpt: “We’re therefore updating the policy to reinforce what has worked well to date, improve the policy where necessary, and implement new measures to incre...”
3anthropic.comCompanies
Anthropic's Responsible Scaling Policy (version 3.0)
Source excerpt: “Responsible Scaling Officer: We will maintain the position of RSO, a designated member of staff who is · responsible for the implementation of this po...”
4anthropic.comCompanies
Anthropic's Responsible Scaling Policy (version 3.0)
Source excerpt: “... We will publish Risk Reports discussing the risks of our systems and how we have made determinations · about whether to continue AI development an...”
5anthropic.comCompanies
Anthropic's Responsible Scaling Policy (version 2.2)
Source excerpt: “Since our founding, we have‬ · ‭recognized the importance ... ‭Usage Policy‬‭sets forth our standards for the use‬‭of our products, including prohibi...”
6aljazeera.comNews and mediaquality 0.78
What came into force with the EU’s AI Act this week – and what didn’t | Technology News | Al Jazeera
Source excerpt: “Published On 6 Aug 20266 Aug 2026 · On August 2, the next phase of Europe’s Artificial Intelligence Act came into force as the European Union frames t...”
7legalnodes.comCompanies
EU AI Act 2026 Updates: Compliance Requirements and Business Risks
Source excerpt: “The EU’s regulatory model for artificial intelligence is founded on a graduated, risk-oriented structure. Rather than targeting specific technologies,...”
8europa.euGovernment
Safer and more transparent AI - European Commission
Source excerpt: “On 2 August 2026, new rules on the transparency of AI systems take effect. AI is advancing quickly, making it increasingly difficult to distinguish AI...”
9softwareimprovementgroup.comCompanies
A comprehensive EU AI Act Summary [August 2026 update] - SIG
Source excerpt: “The Omnibus confirms two fixed dates for the Act’s toughest tier: high-risk obligations for stand-alone systems (Annex III, covering hiring, credit sc...”
10softwareimprovementgroup.comCompanies
A comprehensive EU AI Act Summary [August 2026 update] - SIG
Source excerpt: “Meanwhile, the transparency duties (Article 50) and the AI Office’s enforcement powers over general-purpose AI (GPAI) providers took effect exactly as...”
11techpolicy.pressNews and media
Where State AI Legislation Stands Half Way Into 2026 | TechPolicy.Press
Source excerpt: “By our count, as of July 1, states have enacted 109 AI and 28 data center laws. This is slightly behind the pace of last year, when by the same date s...”
12phillipslytle.comCompanies
Executive Order Issued to Restrict State Regulation of AI
Source excerpt: “Written By: E. Barlow Keener, Lavanya Sathyamurthy · On December 11, 2025, the Trump Administration issued Executive Order 14365, titled, Removing Bar...”
13phillipslytle.comCompanies
Executive Order Issued to Restrict State Regulation of AI
Source excerpt: “Section 3 of the Executive Order requires that, by January 10, 2026, the DOJ establish an “AI Litigation Task Force” to challenge state AI laws deemed...”
14paulhastings.comCompanies
President Trump Signs Executive Order Challenging State AI Laws | Paul Hastings LLP
Source excerpt: “Evaluation of State AI Laws: The Executive Order directs the Secretary of Commerce to publish, by March 11, 2026, a comprehensive review of existing s...”
15paulhastings.comCompanies
President Trump Signs Executive Order Challenging State AI Laws | Paul Hastings LLP
Source excerpt: “We anticipate that the validity of targeted state laws will likely be determined through prolonged litigation that could reach the Supreme Court, wher...”
16mcdermottlaw.comCompanies
New executive order shifts US AI policy toward national security
Source excerpt: “The June 2026 EO presents a new affirmative national security and cybersecurity agenda and introduces considerations that require On June 2, 2026, P...”
17gaeedu.orgNonprofits and think tanks
China AI Governance Profile 2026: Regulation, Policy & Workforce Implications
Source excerpt: “... China’s AI governance model is best understood as a layered regulatory architecture rather than a single comprehensive AI statute. Its foundation ...”
18pertamapartners.comCompanies
China AI Regulations 2026: Rules Companies Must Follow
Source excerpt: “Organizations deploying AI in China face mandatory registration, security reviews, content filtering, data localization, and regular reporting require...”
19twobirds.comCompanies
China's New Regulations on AI Anthropomorphic Interactive Services - Bird & Bird
Source excerpt: “The Measures prohibit seven categories of activities, incorporating general requirements from the Generative AI Measures (safeguarding national securi...”