AI Regulation & Policy

Today's Stories

1. 1

   AI Agent Governance vs. Observability: What's the Difference?

   AI Agent Governance vs. Observability: What's the Difference?

2. 2

   AI investment is shifting from efficiency to innovation and growth, but corporate leaders are losing confidence in their ability to scale AI across their organizations.

   Akkodis' recent report shows that innovation has become the primary driver of AI investment, replacing efficiency as the top focus. At the same time, CTO confidence in scaling AI has dropped significantly due to challenges in integrating AI across enterprise systems. This signals a tension between corporate ambition and execution capability. While companies are betting on AI for new business models and growth, their technical leaders are struggling with the practical work of deploying AI within existing business infrastructure—a gap that could slow the pace of actual AI adoption in enterprises.

   The report identifies agentic AI (AI systems that make decisions and complete tasks with minimal human intervention) as a crucial emerging trend, one that will demand new governance frameworks as these systems become more autonomous.

3. 3

   Trump administration signs orders to accelerate quantum computing development, aiming for a machine capable of scientific research by 2028, as the technology moves closer to commercial use.

   US President Donald Trump signed executive orders to speed up quantum computing development with a target of creating a machine capable of scientific research by 2028. The administration has taken $2 billion（約3200億円） in equity in quantum firms and stakes in other technology companies. Quantum computing is expected to speed drug discovery and materials science, but also poses security risks — Google warned in March that firms should be ready for 'post-quantum cryptography' by 2029. Several quantum firms went public this year, signaling that the previously futuristic technology is moving toward commercialization.

   The 2028 deadline for a quantum machine capable of scientific research, and whether the administration's equity stakes and tech intervention approach will accelerate the timeline compared to private-sector efforts.

4. 4

   I Shot Films for 30 Years. Now I'm Building Safety Systems for AI Agents

   I Shot Films for 30 Years. Now I'm Building Safety Systems for AI Agents

5. 5

   Holden Karnofsky acknowledges AI safety efforts may ultimately have negative impact, citing governance risks and unintended consequences.

   Holden Karnofsky published a list of potential downside risks for AI safety work broadly, noting he takes seriously the possibility that AI safety interventions could cause net harm. He cited AI governance interventions as obviously high-variance, with bad regulation easily making things worse and many interventions potentially increasing the risk of great power conflict. Even leaders in AI safety acknowledge they must live with the possibility that their impact could be negative rather than positive. This reflects a fundamental humility about the complexity of safety work—interventions intended to reduce risk may inadvertently create new risks, particularly in areas like regulation and geopolitics.

   Karnofsky frames this as a reality he must accept while continuing to do his best and hope for the best, suggesting that uncertainty about net impact is an inherent challenge in the field rather than a reason to abandon safety work.

6. 6

   Most impactful AI governance work happens invisibly inside government, not in public debate—and the AI community may be overlooking it.

   A researcher argues that strategic AI governance work occurs largely in two invisible domains: within ministerial cabinets and international forums, and inside national and international institutions. This contrasts with the more visible public work—press releases, open letters, statements—that typically receives attention on platforms like LessWrong. The argument suggests the AI community has a blind spot. By focusing on intellectual production and public-facing work, the field may be underinvesting in the executive-branch and institutional work that actually shapes policy outcomes. This is offered as one reason for hesitations about replicating certain governance initiatives (like ControlAI) across countries, since the conditions for invisible insider work differ by context.

   The researcher identifies a structural bias in how the community recognizes and values different types of governance labor. Understanding which work is truly impactful—rather than which is most visible—may require shifting how the field allocates attention and resources.