OpenAI accuses Meta and Musk of attacks

Key Takeaways

• OpenAI has accused Elon Musk and Meta of coordinated attacks aimed at seizing control of AI development, while Musk seeks $134 billion in damages from OpenAI and Microsoft, requesting funds go to OpenAI's nonprofit foundation.
• Beijing-based Z.AI's GLM-5.1 model leads the SWE-Bench Pro coding evaluation with a score of 58.4, surpassing models like GPT-5.4 and Claude Opus 4.6, while Anthropic's unreleased Claude Mythos Preview scored 77.8% on the same benchmark.
• Cyvl, a Somerville company, uses AI with vehicle-mounted cameras to help over 500 cities, including Boston, efficiently plan road repairs by analyzing road conditions.
• Rochester, Minnesota, is pursuing grant funds for AI-powered automated license plate readers (ALPRs), raising privacy concerns among some lawmakers.
• Missouri lawmakers are delaying AI regulation efforts, including proposals for AI liability, due to fears of losing nearly $900 million in federal broadband funding.
• Raindrop Digital launched The SIGNAL Method, a post-Agile product development framework for the AI era, which replaces two-week sprints with milestone-driven delivery.
• A new Ethereum standard, ERC-8211, is proposed to enable AI agents to execute complex decentralized finance (DeFi) trades in a single, more efficient step.
• The Osceola County School Board approved CGI Advantage, an AI-powered Enterprise Resource Planning system, to modernize its K-12 budgeting, finance, HR, and payroll systems by July 2027.
• AI scribes are increasingly used in doctor's offices for note-taking, with patients having the right to decline their use and needing to review notes for accuracy.
• A writer reflected that while AI-generated email replies are functional, they lack the personal depth and emotional connection found in human communication.

OpenAI Accuses Musk and Meta of Coordinated Attacks

OpenAI has sent a letter to the attorneys general of Delaware and California, alleging that Elon Musk and Meta are working together to attack the company. The letter asks for an investigation into Musk's actions, which OpenAI claims are designed to seize control of artificial intelligence development. Musk is currently suing OpenAI over its shift to a for-profit model. OpenAI believes these attacks aim to put AI control into the hands of competitors without safety principles.

Musk Seeks Damages for OpenAI's Nonprofit Arm

Elon Musk has updated his lawsuit against OpenAI, requesting that any damages he wins go to the company's nonprofit foundation instead of himself. He also wants Sam Altman removed from the nonprofit's board. This is the latest development in the ongoing dispute between Musk and OpenAI. The lawsuit is expected to go to trial soon in California.

OpenAI Urges Investigation into Musk's AI Control Efforts

OpenAI has asked California and Delaware attorneys general to investigate Elon Musk for alleged anti-competitive behavior. OpenAI claims Musk is trying to gain personal control over artificial intelligence development. This request comes just weeks before Musk's trial against OpenAI and Microsoft, where he seeks $134 billion in damages. Musk alleges OpenAI betrayed its founding mission by accepting funding from Microsoft and planning to become a for-profit company.

China's Z.AI Model GLM-5.1 Leads Global Coding Benchmarks

Beijing-based Z.AI has released its GLM-5.1 model, which now leads the world in the SWE-Bench Pro coding evaluation. GLM-5.1 scored 58.4, surpassing models like GPT-5.4 and Claude Opus 4.6. This marks the first time a Chinese model has topped the SWE-Bench Pro leaderboard. The model also shows strong performance in other coding and agentic benchmarks, though US models still lead in pure reasoning tasks.

Anthropic's Claude Mythos Preview Excels in Coding Tests

Anthropic's unreleased Claude Mythos Preview model has achieved top scores on several coding benchmarks, including SWE-Bench Pro with a 77.8% score. It also performed exceptionally well on SWE-bench Verified and SWE-bench Multilingual. Mythos Preview also showed significant improvements in reasoning and agentic tasks compared to the current Claude Opus 4.6 model. Anthropic is restricting access to Mythos Preview due to its advanced cybersecurity capabilities.

Somerville Company Uses AI for Smarter Road Repair Planning

A Somerville, Massachusetts company called Cyvl is using artificial intelligence to help cities manage road repairs more efficiently. Their technology uses vehicle-mounted cameras and sensors to collect data on road conditions. AI then analyzes this data to identify and measure defects, helping cities prioritize and plan repairs. This approach provides cities with better, up-to-date information to manage their infrastructure, with over 500 customers including the City of Boston.

Rochester May Use AI License Plate Readers with Grant Funds

Rochester, Minnesota, and Olmsted County have applied for a grant that could fund the installation of AI-powered automated license plate readers (ALPR). If awarded, the Rochester police department plans to use the funds for this technology. ALPRs use AI to capture vehicle information, but some, like State Representative Greg Davids, express concerns about privacy and the potential for a surveillance state.

AI Email Replies Lack Human Touch, Writer Argues

A writer reflects on their first experience with AI-generated email replies, finding them functional but lacking personal depth. They argue that while AI can provide quick responses, human writers can offer more detail and emotional connection. The author compares AI's role to calculators in math education, suggesting it will be a tool rather than a replacement for essential human communication skills. They believe AI-generated content will not achieve the lasting impact of human-created art.

Missouri AI Regulations Stall Over Broadband Funding Fears

Missouri lawmakers are stalling efforts to regulate artificial intelligence due to concerns about losing federal broadband funds. A Senate bill aimed to establish liability for AI harm, but legislators fear it might violate federal guidelines set by the Trump administration. These guidelines could jeopardize nearly $900 million in remaining funds for rural internet expansion. The bill also proposed rules against AI personhood and marriage. The debate highlights the tension between AI regulation and securing essential infrastructure funding.

New SIGNAL Method Replaces Agile for AI-Era Product Development

Raindrop Digital has launched The SIGNAL Method, a new product development framework designed for the age of AI. This post-Agile approach replaces two-week sprints with milestone-driven delivery and user stories with precise build prompts. The method focuses on product thinking, user understanding, and creating a 'learning machine' that improves over time. Raindrop Digital is also releasing Storm, an AI-powered platform based on The SIGNAL Method to help non-technical founders build products.

Ethereum Proposal ERC-8211 Simplifies AI DeFi Trades

A new Ethereum standard, ERC-8211, aims to allow AI agents to execute complex decentralized finance (DeFi) trades in a single step. Currently, DeFi transactions often require multiple separate actions due to unpredictable outputs like token swap amounts. ERC-8211 enables each transaction step to reference the result of the previous one, making complex financial operations more efficient for AI. This standard works with existing Ethereum infrastructure and enhances usability for AI agents.

Osceola Schools Adopt AI System for Efficiency

The Osceola County School Board has approved the implementation of CGI Advantage, an AI-powered Enterprise Resource Planning system. This system will integrate budgeting, finance, HR, payroll, and reporting into a single cloud platform for K-12 education. District leaders believe this modernization effort, which replaces aging systems, will improve efficiency, transparency, and the employee experience. The project is set to begin April 10, 2026, with a planned go-live date in July 2027.

AI Scribes in Doctor's Offices: Your Rights and What to Know

AI scribes are increasingly used in doctor's offices to help with note-taking, freeing up clinicians to focus on patients. If your doctor uses an AI scribe, they should ask for your permission, and you can request to pause or decline its use. It's important to review the AI-generated notes for errors, as these tools can make mistakes. AI companies may access and use de-identified patient data to improve their software, subject to HIPAA regulations.

Sources

• In Letter, OpenAI Reportedly Says Elon Musk and Meta Are Coordinating 'Attacks' Against It
• Elon Musk Asks for OpenAI’s Nonprofit to Get Any Damages From His Lawsuit
• Elon Musk's $134 billion fight with OpenAI escalates with investigation call
• China’s Z.AI Releases GLM-5.1, Beats All US Models On SWE-Bench Pro
• Anthropic’s Claude Mythos Preview Smashes Coding Benchmarks, Scores 77.8 On SWE-Bench Pro
• Somerville company uses AI to improve cities' handling of road repairs
• Grant could bring artificial intelligence vehicle tracking to Rochester
• First AI glance didn’t measure up to the art of writing
• Missouri AI regulations stall as lawmakers fear loss of rural broadband funds
• Goodbye Agile. Hello SIGNAL. The new way of building and managing products in the Al era.
• Proposed Ethereum Standard Aims to Help AI Agents Execute Complex DeFi Trades
• 'We're built for this': Osceola schools approve AI-powered system
• Can I Opt Out of Having My Doctor Take Notes With AI?