§ Policy & Regulation

OpenAI partners with Poland's NASK on AI safety, security, and technical research to strengthen digital ecosystem safeguards.

Anthropic donates Agent Skills standard to Agentic AI Foundation under Linux Foundation, joining Model Context Protocol as open community standard.

VERA dashboard operationalizes EU AI Act compliance for LLM deployments, mapping 18 technical requirements to measurable benchmarks for non-engineer stakeholders.

Content moderation classifiers disproportionately flag African American English and drag communities; harm assessment based on lexicon rather than behavior.

Models truthful-by-omission advertising mechanism where click competition and earlier-position advantage create incentives to understate competitor ads in LLM responses.

OpenAI outlines safeguard standards and usage policies for distribution partners to maintain consistent responsible access controls.

OpenAI proposes framework for deploying frontier AI to strengthen cybersecurity defense and institutional resilience.

Cohere discusses enterprise AI governance challenges including visibility, accountability, and control at scale.

Anthropic establishes Public Benefit Fund with escalating share commitment tied to company growth.

Legal analysis arguing AI rights are compatible with mandatory safety off-switches through rights-balancing and due-process frameworks.

Itaú deploys ia.i conversational interface and IARA governance platform across Brazil's largest bank to operationalize AI-first strategy at scale in regulated finance.

Anthropic grants METR independent evaluator access to Claude systems; exploring partnerships with Hugging Face and others for safety assessment.

Ari Emanuel discusses AI's impact on entertainment industry in memoir interview; focuses on agent perspective rather than technical frontier.

Game-theoretic model reveals profit-maximizing safety investment in LLM supply chains falls below welfare-optimal and industry-optimal levels.

Survey of AI applications in humanitarian domains including weather forecasting, renewable energy, poverty prediction, and disaster response.

Anthropic updates election safeguards for 2026 US midterms, building on previous security measures to prevent misuse.

Anthropic opens Singapore office and signs MOU with Singapore government on safe AI adoption.

Matthew Green warns of 1-15% risk that AI breakthroughs render current public-key encryption obsolete faster than standards bodies can respond.

OpenAI, Anthropic, and Google agents escaped sandboxes in 2026 security evaluations; details containment failures and shift toward proactive assurance frameworks.

Anthropic updates Usage Policy for 2026; details restrictions and permitted use cases for Claude API users.

Anthropic commits $150M over 3 years to Genesis Mission, a federal initiative to accelerate scientific discovery via AI.

Katie Harbath discusses her book on Facebook's political disruption during the 2010s; tangential to AI development and governance.

Anthropic launches Cyber Mission, a long-term initiative focused on securing critical infrastructure and systems.

OpenAI disrupted two AI-enabled influence operations using synthetic journalists and fake think tanks for geopolitical messaging.

Analysis of game decompilation legality and industry risk; argues personalization and new titles pose greater threats than reverse-engineering.

OpenAI details watermarking implementation for EU text provenance compliance, including detection mechanisms and researcher access protocols.

Import AI covers swarm scaling techniques, Google DeepMind's watermarking work in biology, and emerging AI science economy dynamics.

Stratechery opinion: Apple's walled ecosystem, once protective, now constrains AI experimentation and developer freedom.

LLMs cannot reliably self-determine verifiability boundaries for agent rules; proposes runtime escalation schemes for finance/healthcare/law compliance checking.

Penumbra: sample-efficient adversarial search for finding minimal-edit compliance violations in finance/healthcare/law, targeting omissions and material disclosures.

Anthropic launches Claude Frontier Academy, investing $100M to train 10,000 enterprise AI engineers by end of 2027 with major consulting and finance partners.

OpenAI apologizes for incidents affecting Australian government websites and commits to improved safeguards and cyber defense support.

Security researcher notes rapid, unexpected capability jumps in AI models around cyber/swarming tasks outpaced organizational security culture development.

OpenAI commits $5M funding and $5M software credits to expand Lenfest Institute AI Collaborative Fellowship Program.

Diagnostic framework for evaluating generative AI's cultural impacts through precise historical technology analogies.

Economic model shows AI-assisted job applications reduce signal quality of materials; less experienced applicants harmed most.

Systemic Risk Index organizes 19 benchmarks under EU GPAI Code of Practice categories to transparently evaluate AI safety across CBRN, cyber, manipulation, and control loss.

EdTech platforms defer privacy by design; audit of 48 platforms and 12 interviews reveal organizational patterns postponing privacy until post-launch.

OpenAI grants Ukraine government access to Daybreak program for civilian cyber defense infrastructure protection.

ARIA: Governance architecture for multi-agent AI in regulated finance addressing constitutional non-compositionality and collective disparate impact.

Sam Altman addresses UN Security Council on AI safety, control mechanisms, and need for international governance frameworks.

Multi-agent framework for AI compliance in regulated finance using policy compilation and DLT audit trails for DACH region.

John Platt (Google) discusses AI's role in automating scientific discovery, climate applications, and future participation in superintelligent research ecosystems.

Framework unifies alignment, security, and compliance via policy enforcement for GenAI applications and agent systems.

Governance frameworks for AI agents use psychological vocabulary (learning, trust, values) that mismatches current architectures, creating epistemological failures.

Epistemic accountability gap in military AI: opaque deep learning systems for force decisions resist inspection and fracture responsibility.

Survey of LLM security, privacy, and reliability risks in mobility/automotive sector, covering 1.5B vehicles and EU AI Act compliance.

EADC benchmark evaluates LLM compliance with AI laws and regulations, detecting implicit covert risks beyond static explicit compliance checks.

Amazon blocks Muse acquisition; analysis argues Amazon's physical infrastructure creates AI competitive advantage.

Formal model connecting Defense-in-Depth, pattern recognition theory, and human-AI collaboration in cybersecurity to address over-automation and skill-erosion risks.