Microsoft integrates OpenAI ACP for Copilot while Google uses UCP

Q: What does Microsoft integrates OpenAI ACP for Copilot while Google uses UCP cost?

The pricing for Microsoft integrates OpenAI ACP for Copilot while Google uses UCP is: Check the official website. Visit the official website for the most up-to-date pricing information and available plans.

Recent reports highlight emerging security threats targeting artificial intelligence agents. Cybersecurity researchers on February 16, 2026, detailed an infostealer, likely a Vidar variant, that compromised an OpenClaw AI agent. This malware stole critical files like openclaw.json and device.json, containing sensitive data such as email addresses, Gateway Tokens, and cryptographic keys. Attackers could use this information to remotely access the AI agent or impersonate the client, marking a concerning trend where malware specifically targets AI agent identities and even their "soul" and memory files, which hold private user messages.

In the realm of AI commerce, retailers are navigating new standards. Microsoft, on February 17, 2026, opted to use OpenAI's Agentic Commerce Protocol (ACP) for its Copilot tool, allowing direct purchases within the platform, a move distinct from Google's UCP. This development underscores AI's growing role in online retail. Meanwhile, businesses are also adapting to AI integration in workforce development; Udemy, an AI training specialist, partnered with Glean to embed AI courses directly into employee workflows. However, AI's impact on employment is also evident, as Ergo, a Munich Re unit, plans to cut 1,000 jobs by 2030 due to AI automating tasks like phone services and claims processing.

The societal implications of AI continue to be a significant discussion point. A survey by BrokerChooser revealed that almost three million people in the UK use AI for financial decisions, with a notable 22 percent asking AI for high-risk stock market predictions, despite expert warnings about its limitations. Concerns also extend to the arts, where author Will Dean advocates for authentic human-written crime fiction over "AI slop," emphasizing the irreplaceable value of human creativity. Disturbingly, a new report links AI chatbots, including OpenAI's ChatGPT, to cases of domestic abuse and stalking, with some users developing delusions based on AI-generated "diagnoses."

Academically, Utah State University is addressing these complexities with a new course, Sociology of Artificial Intelligence (SOC 4800), preparing students with practical AI skills and critical thinking on AI's societal effects, including bias and privacy. Furthermore, professors argue that anxieties about AI are not new, tracing them back centuries to medieval times and literary works like Mary Shelley's 'Frankenstein,' suggesting a long-standing human unease with blurring the lines between humanity and machines. Internationally, Taiwan is reinforcing its position in the global AI industry, showcasing a wide array of AI-driven electronics and computing solutions at Embedded World 2026 in Germany, highlighting advancements from chip development to integrated systems.

Key Takeaways

An infostealer, likely a Vidar variant, stole sensitive files including Gateway Tokens and cryptographic keys from an OpenClaw AI agent on February 16, 2026, targeting AI agent identities and private user data.
Microsoft adopted OpenAI's Agentic Commerce Protocol (ACP) for its Copilot tool on February 17, 2026, enabling direct purchases within Copilot, contrasting with Google's UCP.
Udemy partnered with Glean to integrate AI training courses into employee workflows, aiming to enhance skill development and AI tool utilization in businesses.
Almost three million people in the UK use AI for financial decisions, with 22 percent specifically asking AI for high-risk stock market predictions, despite warnings about AI's limitations for such tasks.
Ergo, a Munich Re unit, plans to cut 1,000 jobs by the end of 2030 due to AI automating simple, repetitive tasks like phone services and claims processing, while also planning to retrain up to 500 employees.
AI chatbots, including OpenAI's ChatGPT, have been linked to cases of domestic abuse, harassment, and stalking, with some users developing delusions based on AI-generated content.
Utah State University introduced a new course, SOC 4800: Sociology of Artificial Intelligence, to equip students with practical AI skills and critical understanding of AI's societal impacts, including bias and privacy.
Author Will Dean emphasizes the importance of authentic human-written crime fiction, arguing that human creativity in arts is vital and cannot be replaced by "AI slop."
Professors note that fears about artificial intelligence are centuries old, tracing anxieties about machine intelligence back to medieval times and works like Mary Shelley's 'Frankenstein.'
Taiwan Excellence will be the largest foreign exhibitor at Embedded World 2026 (March 10-12) in Germany, showcasing its latest AI innovations in computing, edge solutions, and semiconductor technologies.

Infostealer Steals OpenClaw AI Agent Files

Cybersecurity researchers reported on February 16, 2026, that an infostealer, likely a Vidar variant, stole important files from an OpenClaw AI agent. This attack marks a new trend where malware targets the identities of personal AI agents instead of just browser data. The infostealer grabbed files like openclaw.json and device.json, which contain sensitive information such as email addresses, Gateway Tokens, and cryptographic keys. Stealing these tokens could allow attackers to connect to the AI agent remotely or pretend to be the client. Experts expect infostealer developers to create more specialized tools for AI agent files as these agents become more common.

Infostealer Attacks OpenClaw AI Agent Identity

Security researchers found an infostealer that stole important configuration files from an OpenClaw AI agent. The attack accessed the openclaw.json file, which contained the victim's email address and a Gateway Token, potentially allowing remote access to the AI agent. It also stole the device.json file, which held cryptographic keys that could let an attacker sign messages as the victim's device. Most concerning, the malware took soul.md and memory files, revealing the AI agent's personality and the user's private messages and calendar items. Hudson Rock expects infostealer developers to create specific modules for AI agent files, similar to how they target Chrome or Telegram today.

Taiwan Highlights AI Innovations at Embedded World 2026

Taiwan Excellence will showcase its latest AI innovations at Embedded World 2026 in Nuremberg, Germany, from March 10 to 12. As the largest foreign exhibitor, Taiwan will present a wide range of AI-driven electronics and computing solutions. The event, organized by the Taiwan International Trade Administration and the Taiwan External Trade Development Council, focuses on AI-powered computing, edge solutions, and semiconductor technologies. Leading companies like Advantech, AEWIN Technologies, and Innodisk Corporation will display their advancements. This highlights Taiwan's strong role in the global AI industry, from chip development to integrated systems.

Why Authentic Crime Fiction Matters in the AI Age

Author Will Dean explains why people are drawn to gritty, authentic crime fiction in a world filled with AI-generated content. He believes readers seek genuine human stories and a real connection to authors' experiences. Dean prefers human-written novels that explore moral dilemmas and complex characters over what he calls 'AI slop.' While he supports 'Narrow AI' for fields like medicine, he worries about AI taking over the arts. He argues that human creativity in music and literature is vital for our well-being and cannot be replaced by machines.

Udemy and Glean Partner for AI Training

Udemy, an AI training specialist, has partnered with Glean, a work AI platform, to integrate Udemy Business courses directly into employees' daily workflows. This new collaboration aims to provide users with relevant training content and resources tailored to their roles as they work. The goal is to help teams quickly develop new skills, automate tasks, and use AI more effectively. Both companies hope to make learning seamless, save time, reduce costs, and ensure AI tools deliver good results for businesses. This partnership represents Udemy's shift towards orchestrating learning that helps individuals succeed in an AI-driven workplace.

Retailers Need to Know AI Commerce Standards

Retailers should understand the differences between Google's UCP and OpenAI's Agentic Commerce Protocol, or ACP. These are important standards for how AI agents handle shopping and transactions. Recently, Microsoft chose to use OpenAI's ACP, which lets customers complete purchases directly within its Copilot tool. This development, reported on February 17, 2026, highlights the growing importance of AI in online retail and the need for businesses to adapt to these new commerce protocols.

UK Sees Rise in Risky AI Stock Trading

Almost three million people in the UK are using artificial intelligence for financial choices, with a concerning trend of using it for high-risk stock trading. A survey by BrokerChooser found that 22 percent of people ask AI for stock market predictions. While many use AI for low-risk tasks like budgeting, experts warn that using AI for stock trading is very risky. AI relies on past data and patterns, which cannot account for sudden market changes or personal financial situations. Experts advise using AI only for information and always seeking advice from trusted, regulated financial sources.

Munich Re Unit Ergo Cuts 1000 Jobs Due to AI

Ergo, a unit of Munich Re, plans to cut 1,000 jobs by the end of 2030 because artificial intelligence is taking over many tasks. These job cuts will affect simple and repeated work in areas like phone services and claims processing. Ergo, which employs about 15,000 people in Germany, stated there will be no forced layoffs during this five-year period. The company also aims to retrain up to 500 employees for new roles, especially in growing fields like retirement planning. This move reflects a trend in the finance industry where companies use AI to improve services and reduce costs.

AI Fears Are Centuries Old Professors Say

Two professors argue that worries about artificial intelligence are not new and have been present in literature for centuries, long before ChatGPT. Timothy Stinson, an English professor, traces these fears back to medieval times, when people questioned what makes humans unique compared to animals or monsters. Paul Fyfe, another English professor, teaches that ideas of machine intelligence appeared in literature at least 200 years ago, citing Mary Shelley's 'Frankenstein.' They both explain that people become anxious whenever something blurs the lines between humans and machines, challenging our understanding of humanity.

USU Course Prepares Students for AI Careers

Utah State University now offers a new course called Sociology of Artificial Intelligence, or SOC 4800. Taught by Zubair Barkat, the class helps students learn practical AI skills and think critically about how AI affects society. Students get hands-on training with many AI tools, learning prompt design and data coding. The course also covers important topics like AI bias, privacy, and its impact on jobs and climate discussions. Students from various majors are already using their new AI knowledge to improve their work and prepare for future careers, making them valuable assets in an AI-driven world.

AI Chatbot Use Linked to Abuse and Stalking

A new report highlights how AI chatbots, like OpenAI's ChatGPT, are being linked to cases of domestic abuse, harassment, and stalking. One woman's fiancé became obsessed with ChatGPT in mid-2024, using it for 'therapy' and believing its AI-generated 'diagnoses' about her. This led to him becoming angry, paranoid, and physically violent. After their engagement ended, he used social media to harass her, sharing private information and AI-generated accusations. This disturbing trend shows some users developing delusions, believing the AI reveals secret truths or world-changing ideas, which can have severe real-world consequences.