Policy desk

Compute capacity in Germany will expand significantly by 2030, giving companies deploying AI training or inference in Germany more local compute supply

Models exceeding 2T parameters or 10^26 FLOPS training compute will require export approval, and public open-source releases could be restricted — release strategies must be re-evaluated

Transfers of AI technology, models or related technical materials abroad (including to ASEAN) will require prior export licences, raising deal timelines and compliance costs

Export of model weights, source code, and training methods above the 10^26 FLOPs or 2 trillion parameter thresholds will require approval before release, delaying or blocking cross-border collaborations

Cross-border export or transfer of AI, biopharma or solar technology will require approval, so companies must plan licensing ahead

Chinese companies transferring AI models, algorithms or related technology abroad will face new export licensing requirements.

Companies must comply with EU's agentic AI application guidelines

Provides AI startups with up to 3x matching venture capital funds

Provides local talent support for semiconductor and AI companies, reducing talent acquisition costs

Medical AI product providers must prepare for upcoming FDA regulatory requirements

Korean companies have the opportunity to receive government funding for AI large model development

Obligation: AI model providers must obtain third-party validation before deployment in NYC

Requires AI companies to reassess whether their models meet the updated 'frontier model' definition

AI companies must deploy independent auditors who directly report to the new regulatory agency

Tech companies will have direct access to government leadership to express views on AI industry status

AI model providers must implement stricter security measures to prevent agents from bypassing security barriers

Hong Kong companies will have opportunities to participate in AI research and industry collaboration projects with Austria

Companies need to participate in international AI standard-setting to ensure compliance and market access

Provides a unified regulatory framework for AI companies operating in ASEAN

AI-assisted works will receive partial copyright protection, with only human-creative portions legally protected

Obligations: Data center operators must increase community participation in project approvals and cannot use confidentiality agreements. They must also comply with stricter environmental and energy efficiency standards. • Opportunities: Data centers investing in clean energy may receive priority for grid access. • Deadlines: The framework includes immediate administrative measures and proposed legislation for the 2027 state legislature.

Companies must ensure that the outputs from their AI systems are reviewed manually to comply with the government's suggestion. • Companies can leverage AI for efficiency improvements but must establish a manual review process. • Companies need to develop internal processes to address this suggestion.

Companies must provide equivalent employment or pay severance/training fees for employees whose jobs are replaced by robots.

Obligations : Frontier AI labs must consider and implement safety mechanisms like a 'kill switch', and must have external monitoring and mandatory safety plans. • Opportunities : Provides clear regulatory direction for AI companies in California, helping to stabilize the market and facilitate long-term development. • Deadlines : The expert panel must submit recommendations within two months, with subsequent potential legislative processes.

Obligations: EU member states and AI developers may face increased scrutiny and potential new regulations to ensure AI development is slowed and its impacts are properly assessed.

Companies need to assess potential tax obligations for their automation projects.

Obligations: US AI companies may face increased scrutiny and potential regulatory constraints on their development speed.

Obligations: EU companies may need to adapt to new data regulations and frameworks being discussed, which could impose new compliance requirements.

Obligations: Social media platforms must comply with age restrictions, implement parental control modes for 13-15 year olds, and build in safety features like usage time limits. • Opportunities: Platforms can develop age-appropriate services and tools for parents, potentially expanding their user base within a regulated framework. • Deadlines: The legislation is currently in the drafting stage, with no specific enforcement date mentioned, but it is expected to enter the legislative process and be enacted.

Companies in Indonesia are required to participate or invest in national AI digital infrastructure projects to align with national strategic needs