Nvidia plans Taiwan headquarters as Google tests Chrome AI

Key Takeaways

• Taiwan's economy is booming due to its role in AI chip manufacturing, with Nvidia planning a new headquarters there and becoming TSMC's largest customer.
• Pakistan pledged $1 billion towards artificial intelligence development and adoption by 2030, announced at its AI Summit on February 12, 2026.
• Cisco expanded its AgenticOps model across networking, security, and observability products, enhancing AI Defense and Cisco SASE features.
• Anthropic CEO Dario Amodei warned that AI will disrupt the job market, potentially affecting software engineering roles rapidly.
• Google's new Auto Browse AI agent in Chrome performed well in some web tasks like playing 2048 but struggled with others, such as creating music playlists.
• Hawaii's Senate Bill 3001, the Artificial Intelligence Disclosure and Safety Act, advanced, requiring AI disclosure and setting safety standards, including protections for minors.
• An 85-year-old woman, Jan Worrell, received an AI robot named ElliQ for free through a pilot program to help her stay in her home.
• A University of Michigan student is suing the university for falsely accusing her of AI use, citing disability bias and unreliable AI comparison tools.
• OWASP released new evaluation criteria for AI red teaming security vendors, emphasizing reproducible testing and transparent metrics.
• OpenAI President Greg Brockman and his wife donated $25 million to a pro-Trump super PAC in September 2025, stating it supports AI development advocacy.

Student Sues University of Michigan Over AI Cheating Claim

A student is suing the University of Michigan, claiming the school wrongly accused her of using AI for class papers. The student, identified as Jane Doe, states her generalized anxiety and obsessive-compulsive disorders affect her writing style, making it appear formal and structured. She argues these traits were misinterpreted as AI use. The university gave her a "no record" grade, which harms her academic future. She filed a lawsuit in federal court in Detroit on February 9, also appealing to the U.S. Department of Education.

Michigan Student Sues University for AI Accusation and Disability Bias

An unnamed University of Michigan student, Jane Doe, is suing the school for falsely accusing her of using AI in her Great Books 191 class. She claims the university denied her disability accommodations during the appeal process, despite knowing about her anxiety and obsessive compulsive disorders since February 2025. The lawsuit states that accusations were based on subjective judgments and unreliable AI comparison tools. Jane Doe seeks damages and a fair appeal, as the "No Record" grade harms her academic standing. Her appeal is currently paused while a civil rights complaint is reviewed.

Taiwan's AI Economy Booms Amid Bubble Fears and China Tensions

Taiwan's economy is experiencing a major boost due to the artificial intelligence boom, driven by its leading role in semiconductor manufacturing. Nvidia plans a new headquarters in Taiwan and will become TSMC's largest customer, further fueling growth. This surge benefits many industries, from logistics to construction, and the government is investing in AI research. However, concerns exist about a potential economic bubble from speculative investments. Ongoing tensions with China also cast a shadow, as any conflict could disrupt Taiwan's vital global chip supply.

Taiwan's AI Boom Drives Economy But Faces Bubble and China Risks

Taiwan's economy is booming, largely due to its strong position in the global AI chip market, with Nvidia planning a new headquarters there. The island's exports surged in 2025, and its economy grew 8.6% last year, supported by companies like TSMC and Foxconn. However, experts like Wu Tsong-min worry about an AI bubble, similar to the dot-com crash, given the heavy reliance on tech. TSMC's chairman C.C. Wei also expressed concern despite massive investments. Additionally, ongoing tensions with China pose a significant geopolitical threat to Taiwan's economic stability.

Cisco Expands AgenticOps Across Network Security and Operations

Cisco announced a major expansion of its AgenticOps model at Cisco Live EMEA in Amsterdam, integrating it across networking, security, and observability products. This agent-driven IT system aims to simplify complex operations and help organizations scale AI effectively. Key updates include AI Defense enhancements like traffic optimization and an AI Bill of Materials for better security. Cisco SASE also gains new features for managing AI traffic and enforcing policies. Additionally, networking products now offer autonomous troubleshooting and improved data center management through AgenticOps.

Anthropic CEO Warns AI Will Disrupt Jobs Rapidly

Dario Amodei, CEO of Anthropic, warns that artificial intelligence will change the world and job market faster than any past revolution. He initially thought entry-level white-collar jobs were most at risk, but now believes software engineering might be disrupted even quicker. Amodei describes a "centaur phase" where AI boosts human productivity, but he worries about the speed of these changes. He questions how people can adapt quickly enough to disruptions happening in just a few years.

Elderly Woman Welcomes AI Robot ElliQ to Stay Home

Jan Worrell, an 85-year-old woman living alone in Washington, received an AI robot named ElliQ in spring 2023. She hopes this companion will help her achieve her goal of staying in her home. A regional nonprofit provided ElliQ for free as part of a pilot program for seniors, covering its $700 annual subscription. The small robot, which looks like a reading lamp, features a tablet and camera and speaks with a cheerful voice.

Hawaii Bill Requires AI Disclosure and Safety Rules

Hawaii's Senate Bill 3001, known as the Artificial Intelligence Disclosure and Safety Act, has advanced through a key committee. This bill sets safety standards for conversational AI services operating in Hawaii. It requires clear disclosure when users are interacting with AI, not humans, and includes stronger protections for minors. The bill also mandates protocols for suicidal ideation and prohibits AI from pretending to be human or a mental health professional. The Attorney General will enforce these new rules.

New Guide Helps Choose AI Red Teaming Security Vendors

Choosing the right vendor for AI red teaming security is becoming more complex, so OWASP released new evaluation criteria. This guide helps organizations assess vendors for both simple and advanced AI systems. It highlights that strong vendors offer reproducible multi-turn evaluations, custom testing, and clear reports linking technical failures to business impact. Buyers should look for transparent metrics like jailbreak success rates and avoid vendors with vague claims or those who only test model output. The guide also emphasizes the importance of operational fit, including CI/CD integration and safe sandboxing.

Chrome's Auto Browse AI Agent Tested on Web Tasks

Google's new Auto Browse AI agent, built into the Chrome browser, was tested on common web tasks. When asked to play the game 2048, it performed well, making 149 moves and creating a 128 tile, despite not using arrow keys. However, it struggled to create a music playlist. The agent refused to monitor a live stream for an hour and failed to add songs to YouTube Music, though it succeeded with Spotify. Overall, the tests showed that while promising, AI agents like Auto Browse are still developing.

Pakistan Pledges 1 Billion Dollars for AI Development by 2030

Pakistan officially launched its Indus AI Week with a major AI Summit in Islamabad on February 12, 2026. During the event, the Prime Minister announced a significant national commitment of 1 billion US dollars towards artificial intelligence development and adoption by 2030. This ambitious pledge shows Pakistan's strategic focus on using AI for economic growth. The week-long event aims to encourage collaboration, display new AI technologies, and attract investments into Pakistan's growing tech sector.

OpenAI President Donates Millions to Trump for AI Future

Greg Brockman, president and cofounder of OpenAI, and his wife Anna Brockman donated 25 million US dollars to a pro-Trump super PAC in September 2025. Brockman states these donations are personal and aim to support OpenAI's mission of developing powerful AI for the benefit of all humanity. He believes AI is the most impactful technology ever created and wants to support leaders who champion its development. While some within OpenAI expressed concern, the company maintains the donations reflect Brockman's personal focus on AI as a critical issue.

Sources

• Student claims U-M wrongly accused her of using AI
• UM student accused of using AI sues school for disability discrimination
• Taiwan's AI-powered economy soars in the shadow of bubble fears and China threats
• Taiwan's AI-powered economy soars in the shadow of bubble fears and China threats
• Cisco extends AgenticOps model across networking, security, observability products
• Video: Opinion | Is A.I. Coming for Your Job?
• To Stay in Her Home, She Let In an A.I. Robot
• Bill advances for disclosure and safeguards on conversational AI services in Hawaiʻi
• Picking an AI red teaming vendor is getting harder
• Tested: How Chrome’s Auto Browse agent handles common web tasks
• Pakistan Opens Indus AI Week With Landmark AI Summit, Announces USD 1 Billion Commitment to Artificial Intelligence by 2030
• OpenAI’s President Gave Millions to Trump. He Says It’s for Humanity