AI Brew

A coalition of companies is advocating for the Chips Security Act (CSA) to enhance security mechanisms for AI chips, preventing them from reaching adversaries like China. The CSA aims to close loopholes in U.S. export controls, ensuring chips are sold only to approved countries. While proponents argue it will boost competitiveness and security, some industry groups fear it could hinder sales. The bill has gained traction in Congress, with unanimous support from the House Foreign Affairs Committee.Regulation

Sen. Bernie Sanders proposed an AI Sovereign Wealth Fund to tax profitable AI companies, allowing citizens to share in their wealth. He also suggested an Independent Commission for Democratic AI to oversee these companies. Sanders and Trump find common ground on giving back to the public, highlighting a growing bipartisan concern over AI regulation and its societal impacts.Regulation

Amazon engineers face internal investigation after criticizing the company's rapid AI data center expansion and advocating for government regulation. Following public testimony, employees reported feeling intimidated by the company's response and alleged monitoring of their political activities.Regulation

Graham Norton won a US court case compelling Meta to disclose the identity behind harmful deepfake posts. The comedian seeks to address false claims affecting his reputation and distressing his family. Meta supports labeling AI-generated content and may ban the offending account under its misinformation guidelines.Regulation

Federal regulators have approved faster connections for large energy users, including AI data centers, to the U.S. electric grid amid rising demand. While tech companies welcome this change, concerns arise over potential impacts on electricity prices and environmental sustainability. The unanimous vote aims to streamline energy access while ensuring that data centers cover their connection costs, addressing community opposition and fears of blackouts as demand grows.Regulation

Google's AI Control Roadmap introduces a comprehensive security framework to safeguard AI agents, treating them as potential insider threats. This strategy enhances traditional cybersecurity measures, ensuring safety even with imperfect alignment. By utilizing trusted AI systems for continuous monitoring and intervention, Google aims to maintain robust defenses as AI capabilities evolve, focusing on risk detection, management, and performance metrics.Regulation

At an OpenAI meeting, employees expressed concerns over donations made to a pro-AI regulation super PAC by a co-founder. Executives sought to distance the company from this PAC, while a new super PAC, Guardrails Alliance, aims to counter such influences with grassroots support for AI regulations.Regulation

The Senate Banking Committee discussed the importance of maintaining American dominance in AI through chip export controls to China. The proposed AI Overwatch Act aims to prevent advanced chip sales to adversaries, while the Match Act seeks to restrict chipmaking tools to China. Both measures are seen as essential for innovation, affordability, and global leadership in AI, as China’s growing capabilities pose a significant threat to U.S. technological superiority.Regulation

India's AI strategy faces challenges as U.S. export controls limit access to foundational models. The country aims to build a sovereign AI stack but lacks the necessary computing power and investment. Recent funding for domestic AI startups highlights the urgency for government support to develop independent AI capabilities.Regulation

The Trump administration imposed export controls on Anthropic's AI technology after concerns over SK Telecom's alleged ties to China. Amazon flagged vulnerabilities in Anthropic's Fable 5 model, prompting the White House to restrict access to US nationals only. Anthropic opted to disable access entirely rather than navigate nationality-based restrictions. The situation escalated after SK Telecom's $100 million investment in Anthropic, raising concerns about security and trust in AI technology management.Regulation

During the G7 Summit, Anthropic CEO Dario Amodei and Google DeepMind CEO Demis Hassabis urged for a U.S.-led coalition to establish global standards for AI, emphasizing international cooperation to mitigate risks associated with powerful AI models. The meeting included tech leaders and G7 heads of state, discussing structured access to AI models and addressing potential threats like cyber and bioterrorism. The call for cooperation comes amid heightened concerns over AI capabilities and recent export controls on new models.Regulation

Anthropic's recent developments highlight the risks of recursive self-improvement in AI, prompting calls for a pause in frontier AI development. The White House's export control ban on Anthropic's models reflects growing concerns over AI's potential for cyber-attacks and the need for regulation to ensure safety.Regulation

China's top securities regulator, Wu Qing, announced a crackdown on technology themes boosting stock prices, particularly those linked to AI. The CSRC will investigate market manipulation and insider trading, reflecting concerns over speculation in the AI stock rally. With the CSI AI index up nearly 30% this year, regulators aim to address financial risks associated with AI tools in trading and prevent companies from misleading investors about their AI capabilities.Regulation

Top AI CEOs, including OpenAI's Sam Altman and Google DeepMind's Demis Hassabis, will meet at the G7 in France to discuss AI risks, infrastructure, and online child protection, highlighting the growing influence of tech leaders in global governance.Regulation

European diplomats aim to collaborate with the U.S. on AI security risks despite recent tensions. Key industry leaders will discuss AI's role in economic growth and societal resilience at the G7 meeting.Regulation

Anthropic's Claude Fable 5 and Mythos 5 AI models were taken offline due to a U.S. export-control directive. The models have advanced capabilities for both cybersecurity and exploitation, raising national security concerns. Experts warn that similar AI capabilities will emerge from other companies soon, highlighting the need for broader regulatory strategies to manage AI advancements in cybersecurity.Regulation

Anthropic has temporarily disabled access to its Fable 5 and Mythos 5 AI models to comply with a U.S. government order. Prediction markets suggest a strong likelihood that access to the Fable model will be restored to U.S. consumers by mid-July, with a 58% chance by July 1 and a 74% chance by July 10.Regulation

Google has introduced a new 'Search Services History' setting that saves users' uploaded images, audio, and video to improve AI models. Privacy experts advise users to opt out, raising concerns about the implications of storing sensitive personal information. This feature, which tailors user experiences and recommendations, has sparked debate over privacy and the potential misuse of biometric data. Users can manage their settings to control what Google tracks and stores.Regulation

A haunted house site in Pennsylvania is facing backlash over a proposed datacenter complex, raising concerns about air pollution and noise for nearby veterans. Local officials rejected the plan amid community opposition, while the governor pushes for AI infrastructure growth, complicating the regulatory landscape. Residents demand a moratorium on new datacenters to assess environmental impacts, as national trends show growing public opposition to such facilities. The conflict highlights the tension between economic development and community health.Regulation

Steam Next Fest showcases 8700 games, with 1704 featuring generative AI, making up 20% of participants. Steam mandates disclosure of AI use, causing controversy among developers and gamers.Regulation