OpenAI launches EVMbench while Pentagon reviews Anthropic

Key Takeaways

• A $15 billion AI data center campus in Wisconsin is causing public concern over land acquisition and potential increases in electricity bills, with a Politico survey showing opposition if bills rise by $25 or more.
• OpenAI and crypto investor Paradigm launched EVMbench, an AI agent tool designed to enhance smart contract security by identifying and fixing vulnerabilities.
• The Pentagon is considering ending its partnership with Anthropic due to ethical concerns regarding AI's potential use as a weapon.
• Tavus introduced Phoenix-4, an AI model capable of creating realistic, emotionally intelligent video with sub-600ms response times and customizable emotions.
• MindBio Therapeutics is developing an Edge AI platform to detect drug and alcohol intoxication via voice analysis, with deployment in Chilean mining planned for Q2 2026.
• The Minnesota State Capitol has implemented new AI security checkpoints to enhance safety for visitors and lawmakers.
• The "Vibecoding" trend, utilizing tools like Claude Code, is emerging for generating software, indicating a shift in software development.
• Indian companies are rapidly integrating AI into leadership, hiring over 50 Chief AI Officers in the past year across diverse industries.
• AI data centers are adopting optical networking equipment to reduce energy consumption and accelerate operations, addressing the high energy demands of AI.
• A report by energy researcher Ketan Joshi challenges Big Tech's claims about AI saving the planet, noting little evidence and the industry's substantial energy consumption, sometimes relying on coal plants.

Wisconsin residents face losing land for AI data center

Tom Uttech and other Wisconsin residents face losing their land due to a new $15 billion AI data center campus. This massive 600-acre facility in Port Washington needs huge power lines, threatening properties like Uttech's 52-acre home and art studio through eminent domain. The Trump administration supports this rapid AI expansion for national security reasons. Port Washington Mayor Ted Neitzke believes the project will bring thousands of jobs and significant tax revenue. Experts like Ari Peskoe from Harvard warn that these data centers use as much electricity as a large city, requiring extensive new infrastructure.

Americans worry about rising bills from AI data centers

Americans are pushing back against the rapid expansion of AI data centers across the U.S. These large facilities demand huge amounts of land, water, and electricity, causing worries about higher electricity bills. A Politico survey shows voters oppose data centers if bills rise by $25 or more each month. Other concerns include blackouts, taxpayer costs, job losses, and environmental damage. Lawmakers are introducing bills to ban construction or make tech companies pay for utility increases, while the Trump administration supports AI development but wants the industry to address public concerns.

AI data centers spark worries over rising electricity costs

The tech industry's rapid building of AI data centers across the U.S. faces strong opposition. These large facilities demand vast amounts of land, water, and electricity, leading to concerns about higher utility bills for Americans. A Politico survey found that most voters would oppose a data center if it increased their monthly electric bill by $25 or more. People also worry about blackouts, taxpayer expenses, job losses, and environmental effects. State lawmakers are proposing temporary bans or requiring tech companies to cover increased utility costs, even as the Trump administration promotes AI development.

OpenAI and Paradigm launch AI tool for smart contract security

OpenAI and crypto investor Paradigm teamed up to create EVMbench, a new AI agent tool. This tool aims to boost the security of smart contracts, which hold billions of dollars in assets. EVMbench tests how well AI agents can find, fix, and even exploit weaknesses in smart contracts. The partnership comes after recent attacks on platforms like Moonwell and CrossCurve. EVMbench uses 120 vulnerabilities from 40 audits to help protect against future cyber threats in decentralized finance.

OpenAI and Paradigm release AI tool for blockchain safety

OpenAI, with help from Paradigm, has introduced EVMbench, a new benchmarking tool. EVMbench works to make blockchain security better by testing AI agents. It measures how well these AI agents can find, fix, and use weaknesses in smart contracts. The tool uses 120 known vulnerabilities from over 40 audits, including scenarios from Tempo L1. This collaboration highlights how artificial intelligence is becoming more important for keeping decentralized finance safe.

Minnesota Capitol uses new AI security checkpoints

The Minnesota State Capitol now uses new AI security checkpoints. These checkpoints offer a new sense of security for visitors and lawmakers. Jeff Wagner explained how this technology works in a segment called "Good Question." Many people may have already passed through similar checkpoints elsewhere.

Pentagon may end Anthropic partnership over AI weapon fears

The Pentagon is considering ending its partnership with the technology company Anthropic. This threat comes from worries about how artificial intelligence could be used as a weapon. Fox News chief national security correspondent Jennifer Griffin reported on this situation. The concerns highlight the growing debate around the ethical use of AI in defense.

Tavus unveils Phoenix-4 for realistic emotional AI video

Tavus has launched Phoenix-4, a new AI model designed to create realistic, emotionally intelligent video for conversations. This model aims to make AI avatars feel more human by overcoming stiff movements and lack of emotion. Phoenix-4 uses a three-part system: Raven-1 for understanding emotions, Sparrow-1 for natural timing, and Phoenix-4 for real-time video rendering using Gaussian-diffusion. It achieves very fast responses, under 600 milliseconds, by streaming video directly. Developers can also control the AI's emotions, like joy or sadness, and create custom digital twins called Replicas from just two minutes of video.

Indian companies hire many new Chief AI Officers

Artificial intelligence is now a major focus for Indian companies, moving into top leadership positions. Over the past year, more than 50 senior AI-focused leaders, including Chief AI Officers, have been hired. These new roles are appearing across various industries such as banking, manufacturing, and media. This trend shows that AI is no longer just for small projects but is becoming central to business strategy in India.

New AI tool Claude Code can write software

A new AI trend called "Vibecoding" uses tools like Claude Code to generate software. This technology can create code for websites and apps. Kevin Roose, a technology columnist for The New York Times, explored this process. This development could greatly change the software development industry.

MindBio develops AI voice tool to detect intoxication

MindBio Therapeutics is creating an Edge AI platform to detect drug and alcohol intoxication using voice analysis. This new hardware and software system will allow real-time testing directly on devices, even in remote areas. MindBio plans to first deploy this technology in the mining sector in Chile, where substance use among workers is a concern. The company expects to begin manufacturing hardware and field deployment in Q2 2026. This scalable solution aims to provide a non-invasive and cost-effective way to improve safety in industries like mining, aviation, and heavy industrial sectors.

Big Tech lacks proof AI will save the planet

A new report by energy researcher Ketan Joshi questions Big Tech's claims that generative AI will help save the planet. Joshi found little proof behind assertions that AI can significantly cut global emissions. The rapid expansion of AI data centers demands huge amounts of energy, sometimes keeping coal plants running. Critics argue that tech companies often mix less energy-intensive AI with high-energy generative AI when making environmental claims. Experts advise caution and careful examination of these statements, as many lack strong evidence.

AI data centers use optical tech to save energy

AI data center companies are now using optical networking equipment inside their facilities. This change aims to cut down on energy use and make operations faster. The soaring demand for energy from artificial intelligence is driving this new interest in optical technology. Many new companies are working to create these optical solutions, and investors are supporting their efforts.

Sources

• A 600-acre AI data center could cost some Wisconsin residents their land
• Data center backlash driven over concerns about increases to electricity bills
• Data center backlash driven over concerns about increases to electricity bills
• OpenAI and Paradigm partner on AI agent tool for smart contract security
• OpenAI Unveils AI Benchmark Tool to Enhance Blockchain Security
• How do AI security checkpoints work?
• Pentagon threatens to end partnership with tech company amid concerns over AI weaponization
• Tavus Launches Phoenix-4: A Gaussian-Diffusion Model Bringing Real-Time Emotional Intelligence And Sub-600ms Latency To Generative Video AI
• AI in Indian companies: Your next boss might be a chief AI officer.
• Can A.I. Already Do Your Job?
• MindBio Advances Large-Scale Deployment of Voice Intoxication Detection Through Integrated Edge AI Hardware-Software Platform
• Big Tech Says Generative AI Will Save the Planet. It Doesn't Offer Much Proof
• AI data center firms are adopting optical gear inside facilities to cut energy and boost speed