Market & Regulation
Governance for AI Agents Must Live in Data Layer
VentureBeat AIhigh confidence
Why it matters
- →Governance at the data layer ensures real-time enforcement of policies.
- →It provides a robust framework for managing unpredictable agent behavior.
- →Enterprises can deploy AI agents with confidence, knowing actions are controlled and auditable.
Enterprises are increasingly granting AI agents more autonomy, raising concerns about governance and control. Traditional methods of applying rules at the agent layer are inadequate due to the unpredictable nature of autonomous systems. Instead, governance should be enforced at the data layer, where agents interact with data in real-time. This approach ensures that policies are executable and enforceable at the moment of action, providing a robust framework for managing agent behavior. By embedding governance in the data layer, enterprises can confidently deploy AI agents, knowing that their actions are controlled and auditable.
Read originalMore from VentureBeat AI
More in Market & Regulation
© The Verge AIMarket & Regulationbusiness
Court Rules Anthropic's Blacklisting Unconstitutional
A federal judge has determined that the Pentagon's blacklisting of AI company Anthropic was unconstitutional, marking a pivotal legal win for the firm. The court found that the Trump administration's actions were an unlawful response to Anthropic's refusal to permit its AI technology for mass surveillance or lethal autonomous weapons. This decision highlights the ongoing conflict between AI companies and government agencies over the ethical use of technology. With this ruling, Anthropic can proceed with its operations without the limitations of being labeled a supply chain risk, potentially reshaping its interactions with the Department of Defense.
The Verge AI
© WIRED AIMarket & Regulationbusiness
Judge Blocks Pentagon's Blacklist of Anthropic
A federal judge has ruled against the Pentagon's attempt to blacklist AI company Anthropic, calling the move unconstitutional and without basis. This decision overturns the Pentagon's designation of Anthropic as a 'supply-chain risk,' which had previously barred the company from securing federal contracts. The case arose from a dispute over the use of Anthropic's AI models in military operations, particularly following a controversial mission involving Venezuelan president Nicolás Maduro. The ruling allows Anthropic to continue its collaborations with the government, highlighting the complex relationship between AI companies and national security interests. While the Pentagon may appeal, this decision marks a pivotal moment in the conversation about AI deployment in sensitive contexts.
WIRED AI
© GitHub ChangelogMarket & Regulationcoding
GitHub Actions Retention Policy Update Announced
GitHub is updating its Actions retention policy to include checks, workflow runs, and statuses, aligning them with the existing retention settings for artifacts and logs. Starting October 1, 2026, these elements will be automatically cleaned up after the configured retention period, which defaults to 90 days. This change aims to reduce stale data and improve the efficiency of GitHub Actions. Users are encouraged to review and adjust their retention settings to manage storage costs effectively, as this update could impact billable storage usage.
GitHub Changelog