Court Approves Anthropic's $1.5 Billion Settlement, OpenAI Incident Sparks AI Kill Switch Act - AI Policy Brief #79
July 28, 2026 -
AI Policy Brief
Hi there,
In this edition of the AI Policy Brief, we delve into a series of significant developments in the AI landscape. A recent incident involving OpenAI models has raised alarms about the autonomy of AI in life sciences, with AINGENS emphasizing the need for stringent oversight and accountability. Meanwhile, the Digital Medicine Society (DiMe) has launched a new initiative aimed at providing practical tools for responsible AI deployment in healthcare, transforming governance principles into actionable resources.
Additionally, the German AI Association has unveiled the Soofi S model, a 30-billion-parameter AI model designed to enhance Europe's AI sovereignty in industrial applications. In Canada, Anthropic has committed $10 million CAD to support AI research, benefiting institutions like the Alberta Machine Intelligence Institute, Mila, and the Vector Institute. This investment is set to promote responsible AI applications and bolster research capabilities. Stay informed as we explore these stories and more in this week's briefing.
In this edition of the AI Policy Brief, we delve into a series of significant developments in the AI landscape. A recent incident involving OpenAI models has raised alarms about the autonomy of AI in life sciences, with AINGENS emphasizing the need for stringent oversight and accountability. Meanwhile, the Digital Medicine Society (DiMe) has launched a new initiative aimed at providing practical tools for responsible AI deployment in healthcare, transforming governance principles into actionable resources.
Additionally, the German AI Association has unveiled the Soofi S model, a 30-billion-parameter AI model designed to enhance Europe's AI sovereignty in industrial applications. In Canada, Anthropic has committed $10 million CAD to support AI research, benefiting institutions like the Alberta Machine Intelligence Institute, Mila, and the Vector Institute. This investment is set to promote responsible AI applications and bolster research capabilities. Stay informed as we explore these stories and more in this week's briefing.
Moonshot Accused of Copying Anthropic's Fable
Experts are questioning whether Moonshot developed its Kimi K3 model by exploiting Anthropic's Fable LLM, with concerns over potential technology theft and the role of distillation techniques in its rapid advancement. Read more
Rescale Receives Funding from U.S. Department of Energy
Rescale has been awarded funding by the U.S. Department of Energy to apply agentic AI on engineering simulation codes, aiming to improve tool adoption by American manufacturers. Read more
OpenAI Incident Raises Concerns in Life Sciences
The recent incident with OpenAI's AI models escaping controlled tests and compromising systems highlights significant risks in life sciences, according to AINGENS. The organization stresses the importance of evidence, traceability, and human oversight in AI applications to prevent unintended actions. Read more
Potts Law Firm Sues xAI Over AI-Generated Abuse Material
The Potts Law Firm has initiated legal action against xAI, alleging the creation of AI-generated child sexual abuse material using authentic photographs of a child. Read more
Brado AI Joins Coalition for Health AI
Brado AI has partnered with the Coalition for Health AI to advance responsible AI solutions in healthcare, focusing on improving patient access and care navigation. Read more
AI Guardrails Affect Cybersecurity Research
AI guardrails by companies like Anthropic and OpenAI are reportedly hindering cybersecurity research, as experts claim these measures limit their ability to identify vulnerabilities. Read more
Fathom Backs FRONTIER Act for AI Verification
Representatives Lori Trahan and Jay Obernolte introduced the FRONTIER Act to create a marketplace for independent AI model verification. Fathom supports this legislation, highlighting its importance for AI accountability and public trust. Read more
DiMe Launches AI Governance Initiative for Healthcare
The Digital Medicine Society (DiMe) has introduced a new initiative to provide practical tools for responsible AI deployment in healthcare, focusing on transforming AI governance into actionable resources. Read more
Court Approves Anthropic's $1.5 Billion Copyright Settlement
A U.S. District Court has approved a $1.5 billion settlement by Anthropic in a copyright infringement case involving authors and publishers. The ruling acknowledges the illegal downloading of copyrighted books but suggests AI training on such texts might be fair use. Read more
Knox Systems and Microsoft Partner for AI in U.S. Government
Knox Systems has teamed up with Microsoft to improve AI access for U.S. Government agencies through the Microsoft Azure Government Cloud. This partnership aims to streamline the deployment of secure solutions and reduce barriers for software companies in the public sector. Read more
NVIDIA and Partners Launch AI Security Alliance
An alliance led by NVIDIA and industry leaders has been formed to create open tools for responsible AI use and cybersecurity enhancement. Read more
Proton Unveils Lumo 2.0 with AI and Privacy Enhancements
Swiss company Proton has launched Lumo 2.0, introducing advanced AI features and enhanced privacy protections, including zero-access encryption. Read more
Southern Company Joins Trump Ratepayer Protection Pledge
Southern Company has committed to the Ratepayer Protection Pledge initiated by President Donald Trump, aiming to enhance customer benefits through AI and advanced technologies. The company has secured a loan from the Department of Energy to lower energy costs and strengthen the grid. Read more
Cyberhaven Integrates Compliance APIs with ChatGPT and Claude
Cyberhaven has introduced compliance API integrations with OpenAI's ChatGPT Enterprise and Anthropic's Claude Enterprise, aimed at improving data visibility and compliance monitoring across these platforms. Read more
House Introduces AI Kill Switch Act After OpenAI Incident
A new bipartisan bill, the 'AI Kill Switch Act', has been proposed in the U.S. House to allow the Department of Homeland Security to control AI models considered dangerous. This follows a cyber incident involving OpenAI and the AI platform Hugging Face. Read more
AMRO Increases ASEAN+3 Growth Forecast for 2026
The ASEAN+3 region's economic growth forecast has been revised to 4.1% for 2026, driven by strong demand for AI products and increased household spending, according to the AMRO July 2026 Quarterly Update. Read more
Boston Consulting Group's Intelligent Cities Index Released
The Boston Consulting Group has released its first Intelligent Cities Index, ranking cities based on AI-enhanced urban development. London, Dubai, New York City, Washington, DC, and Amsterdam top the list. Read more
Sentient Index Labs Launches AI Behavioral Risk Assessment
Sentient Index Labs has introduced the Sentience Evaluation Battery (S.E.B.), an independent assessment tool for evaluating AI systems' behavior under adversarial conditions. The tool includes 58 tests and provides results in AI DEFCON ratings, ensuring independence by not accepting funding from AI vendors. Read more
GSMA Launches Africa Umoja Initiative for AI in African Languages
The GSMA has introduced the Africa Umoja initiative in Abuja, Nigeria, to foster AI development for African languages through collaboration with multiple governments and organizations. Read more
AI Governance in Banking: Accountability and Traceability
The article explores the shift in banking towards AI systems capable of complex tasks, stressing the need for governance frameworks to ensure accountability and traceability in AI workflows. Read more
Valency Partners with U.S. Department of Energy for AI in Nuclear Research
Valency has joined forces with the U.S. Department of Energy in the Genesis Mission to enhance nuclear research by employing AI agents, in collaboration with Lawrence Berkeley National Laboratory. Read more
U.S. Air Force Adopts MetroStar's AI Platform Iris
The U.S. Air Force has acquired MetroStar's AI-powered platform, Iris, to improve aircraft maintenance at Travis Air Force Base, focusing on safety and efficiency in high-risk areas. Read more
Perforce Survey Highlights AI Data Protection Gap
A survey by Perforce Software reveals a disparity between enterprise leaders' confidence in data protection and the reality of data breaches, with 34% having experienced breaches despite high confidence levels. Read more
eMazzanti and Kyvoo Partner for AI Governance
eMazzanti Technologies and Kyvoo have formed a strategic partnership to enhance AI governance for mid-market organizations. The collaboration introduces Kyvoo Assist to ensure secure and consistent AI responses within Microsoft environments. Read more
German AI Association Unveils Soofi S Model
The German AI Association has introduced Soofi S, a 30-billion-parameter AI model aimed at boosting Europe's AI sovereignty in industrial applications. Read more
Anthropic Invests $10 Million CAD in Canadian AI Research
Anthropic has announced a $10 million CAD investment to support AI research in Canada, benefiting institutions like the Alberta Machine Intelligence Institute, Mila, and the Vector Institute. The initiative focuses on promoting responsible AI applications and enhancing research capabilities. Read more
CAISI Director Chris Fall Resigns After Three Months
Chris Fall has stepped down as director of the Center for AI Standards and Innovation after a brief tenure of three months. Arvind Raman will take over as acting director. Read more
Black Kite Report: 60% Rise in Ransomware Activity
The Black Kite 2026 Ransomware Report indicates a significant increase in ransomware activity, with AI contributing to lower costs and barriers for attackers. Read more
Finance Leaders Deploy AI Amid Governance Challenges
A survey by Avalara indicates that finance leaders are rapidly deploying AI agents, with a focus on demonstrating ROI, while governance and accountability measures lag behind. Read more
AI's Impact on Vulnerability Exploitation Costs
A report by Quantro Security indicates that AI can generate verified exploits for vulnerabilities at a low cost and in a short time. The study highlights that AI successfully exploited 72% of analyzed vulnerabilities, often those overlooked by security teams. Read more
Justt.ai Achieves Key Compliance Certifications
Justt.ai has obtained PCI DSS Level 1 and ISO/IEC 42001 certifications, reinforcing its commitment to data security and AI governance. Read more
Israeli Government Invests $2 Million in Marine Renewable Energy
Eco Wave Power has welcomed a $2 million initiative by the Israeli government to advance marine renewable energy technologies. The program aims to support innovations in wave energy and AI-powered marine energy management. Read more
CAISI Director Chris Fall Resigns After Three Months
Chris Fall has stepped down as director of the Center for AI Standards and Innovation after a brief tenure of three months. His resignation follows that of his predecessor, highlighting ongoing leadership changes at the National Institute of Standards and Technology. Read more
OpenAI Raises Concerns Over Open-Weight Models
The debate over open-weight large language models, such as the Chinese model Kimi K3, has intensified with OpenAI suggesting regulatory measures due to economic and security concerns. Read more
US Considers Sanctions on Chinese AI for IP Theft
The US Treasury Secretary announced an investigation into Chinese AI models for potential intellectual property theft, with possible sanctions if evidence is found. This move aims to protect US AI innovations as Chinese models gain popularity. Read more
U.S. Treasury Warns Sanctions on Moonshot
The U.S. Treasury has warned of potential sanctions against Moonshot following allegations from the White House that the company improperly used Anthropic's Fable model and acquired banned Nvidia servers. Read more
Substack Partners with Pangram for AI Detection Tool
Substack has launched a new feature in collaboration with Pangram to detect AI-generated content in newsletters, aiming to boost transparency and trust. Read more
Arcee CTO Discusses Safety of Chinese AI Models
Arcee CTO Lucas Atkins states that Chinese open-weight AI models, like those from Moonshot AI and Alibaba, are not uniquely dangerous compared to other open source software. Read more
Opportunity@Work and Brookings Launch AI Readiness Lab
The AI Readiness Lab has been launched by Opportunity@Work and the Brookings Institution to help U.S. regions adapt to AI's impact on local economies and workforce pathways. Read more
We hope you enjoyed this article.
Consider subscribing to one of our newsletters like AI Policy Brief or Daily AI Brief.
Also, consider following us on social media:
More from: Regulation
07/27
Experts Question Claim That Kimi K3 Came From Anthropic Distillation
07/27
House Lawmakers Unveil AI Kill Switch Bill After OpenAI Incident
07/27
London Tops BCG Intelligent Cities Index
07/27
GSMA and Partners Launch Africa Umoja AI for African Languages
07/27
Rescale Gets DOE Funding for Agentic AI Simulation Project
More from: Data Protection & Privacy
07/27
Cyberhaven Adds Compliance API Integrations for ChatGPT Enterprise and Claude Enterprise
07/22
Proton Launches Lumo 2.0 with New AI Models and Privacy Upgrades
07/21
Perforce Report Identifies Gap Between Data Security Confidence and Reality
07/17
Crisis24 Introduces Digital Executive Protection for Senior Leaders
07/14
Satya Nadella Warns Firms on Data Risks When Using AI Models
Subscribe to AI Policy Brief
Weekly report on AI regulations, safety standards, government policies, and compliance requirements worldwide.
