Microsoft Saves $500 Million Thanks to AI

According to a Bloomberg News report, Microsoft has saved over $500 million in its service centers in the past year due to the use of artificial intelligence1. It also announced the layoff of an additional 9,000 employees (in addition to 6,000 layoffs in previous months). This is part of an effort to expand data centers to deal with bottlenecks, reduce costs, and invest huge sums in AI infrastructure.

AI Agents Are Integrating into Organizations

Walmart announced the launch of AI agents that will handle customer service, inventory management, supplier support, and software development, with the goal of strengthening its digital commerce sector5. The project demonstrates a trend of AI implementation and integration with organizational systems to improve the shopping experience and increase efficiency.

Cause for Concern or a Wake-Up Call?

Salesforce CEO Marc Benioff said that artificial intelligence successfully performs between 30% and 50% of the work at the company, and the company is considering whether it is even worthwhile to continue hiring human engineers. On “The Joe Rogan Experience” podcast, Meta CEO Mark Zuckerberg said that this year alone, artificial intelligence will replace mid-level engineers at the company. They are not alone; many other large companies have previously written about becoming more efficient with AI and cutting staff.
It appears that all the declarations about replacing people with AI are intended to impress investors, maintain stock value, and justify cuts that are ultimately aimed at increasing profits. While these statements may not be entirely accurate, AI technology is developing at an exponential rate, and organizations are starting to implement AI in their workflows. I receive many requests from C-Level managers in organizations asking for help with AI implementation to reduce employees. So, don’t let them put you to sleep!

Managers Use AI to Fire Employees

A Resume Builder report published some interesting data:

• 65% of managers in the US report using AI at work.
• 94% of them use it to make decisions about their employees.
• About 78% rely on AI for determining salary raises, 77% for promotion decisions, and 66% for termination considerations.
• 21% of managers let AI make final decisions without human intervention.
• Two-thirds of managers who use AI have not received formal training on ethics or how to integrate these tools into employee management.
• Nearly half of the managers who were tasked with investigating if employees could be replaced by AI actually performed a full replacement of positions with AI.

One of the main challenges organizations need to deal with is how to integrate artificial intelligence responsibly into management processes, so that it serves as a recommendation engine while leaving the final decisions to the managers.

Unprecedented Fraud Crisis

On one hand, artificial intelligence helps us become more efficient and reach life-saving solutions much faster, and on the other hand, irresponsible use of it brings many risks. OpenAI CEO Sam Altman warns of a “significant fraud crisis” that will soon hit the world as a result of the accelerated development of artificial intelligence. The ability to easily forge voices, images, and videos that are difficult or impossible to distinguish from reality poses a particular threat in the financial sector, where voice authentication is no longer secure and other solutions need to be considered.

Anti-Deepfake Law

Denmark is leading a first-of-its-kind initiative in Europe to protect personal and creator identity in the age of AI deepfakes, as it will be difficult to distinguish between real and fake content. Denmark is proposing an amendment to its copyright law with the aim of giving all its citizens legal ownership of their face, voice, and body to combat deepfakes created by artificial intelligence. The law is intended to protect both the general public from realistic digital forgery of personal characteristics without consent, and artists from unauthorized imitations of their work by artificial intelligence. The law imposes severe sanctions on technology platforms that do not act to remove content, in accordance with the requirements of the EU’s Digital Services Act (DSA). The bill will be submitted for approval by the European Commission, will be adapted to European human rights conventions, and will preserve the freedom of satire and artistic expression. It is expected to pass in the Danish parliament by the end of this year or early next year.

Indonesia, Taiwan, and China

In the coming month, Indonesia will complete a national program that serves as a clear message to foreign investors, aiming to attract investment and join the global race for AI and the chip industry.

The Taiwanese government launched ten AI infrastructure initiatives with the goal of generating an estimated $510 billion by 2040, including investments in silicon photonics, quantum computing, and robotics. The plan will include the establishment of three global research labs and the creation of about half a million jobs.

At a conference in Shanghai, over 800 companies, including Chinese giants Huawei and Alibaba, showcased thousands of notable innovations in the field of artificial intelligence despite US restrictions on chip exports. The exhibition featured smart robots, advanced chips, smart AI screens, augmented reality glasses for education and shopping, developments in digital health, innovations in digital commerce, and renewable energy, among others.

Chinese Premier Li Qiang, at the WAIC conference, proposed the establishment of a global organization to manage the development of artificial intelligence and enable broader regulatory cooperation. The goal is to prevent concentration among major powers or single companies and expand access to additional countries.

Rules for AI Use in Courts

The California court system issued an official directive in mid-July 2025, stating that all courts in the state must clearly prohibit or regulate the use of generative artificial intelligence (generative AI) by judges and court staff by early September. This is the first decision of its kind in the United States, and it was drafted by a special team established last year by the Chief Justice of the California Supreme Court. The new rules allow each court to choose between two options:

• A complete prohibition on the use of generative AI.
• The formulation of a policy that clearly regulates the use of artificial intelligence tools, including policies for protecting confidential information, preventing AI discrimination and bias, requiring human verification of AI-produced materials, and reporting on documents created using AI. The purpose of the directive is to balance technological innovation with the protection of privacy, fairness, and the public, as well as to prevent safety risks, data security issues, and legal errors. All policies must come into effect by September 1, 2025. Courts that approve the use of AI will be required to implement the rules of use by December 15, 2025, thereby ensuring ongoing oversight, accountability, and transparency.

Turkey Blocked Grok

A Turkish court has blocked access to the Grok chatbot from Elon Musk’s X platform. The reason given was that the platform did not activate sufficient filtering mechanisms, which could expose users to false or inciting information. In practice, Grok was blocked after it was found to contain responses in Turkish that were considered offensive to President Erdoğan, who did not approve of content that could be interpreted as politically inciting or biased. This is the first documented case where a country has blocked an artificial intelligence tool for insulting a political leader.

Innovations in ChatGPT

OpenAI has added a new tool to ChatGPT called “Collaborative Learning” that will make learning interactive. Instead of giving you the correct answer, it will guide you on how to solve problems yourself and help you master new concepts. OpenAI also released the ChatGPT Agent, a virtual agent that performs actions on your computer. It searches for information on websites, clicks buttons, opens applications, fills out forms, places orders, creates presentations, and can perform many other tasks. What it essentially does is perform multi-step processes through smart automation and under the user’s full control. While it’s important to note that it’s still not working perfectly, this is a significant step that leads to the next stage in the evolution of artificial intelligence applications. If you are a Pro subscriber and have a Mac, you can use the option to record calls/meetings, get real-time transcriptions, and summarize and extract important insights. OpenAI announced that it will launch GPT-5 in the coming days, the next generation of its language models, with the potential for a significant upgrade in conversational, understanding, and creative abilities. OpenAI is also launching an AI-based browser with the goal of replacing Google Chrome. The new browser will also perform actions like placing orders and filling out forms.

OpenAI, together with Microsoft, Anthropic, and the American Federation of Teachers (AFT), is establishing a national academy for teaching with artificial intelligence, which will provide training to 400,000 teachers in the US.

Monday.com launched the Vibe Coding tool, which allows anyone to develop applications without coding knowledge. It also launched Sidekick, a bot based on organizational information that can provide answers based on that data, and the Magic tool for creating additional features using only prompts.

An open-source model called ThinkSound can analyze video (images, movements, scenes) and add a smart and accurate soundtrack to every frame. You can try it out at this link.

Anthropic launched a new learning platform where you can take free courses on Claude. Upon completing the course, you can also receive a certificate. It also opened up the possibility of connecting to a variety of tools like Canva, Notion, PayPal, and more.

Runway launched Act-Two, which identifies and mimics body, head, face, and hand movements. It also launched Aleph, which can edit a video, including adding and deleting objects, changing lighting, camera angle, and background, and anything else you want to change in the video.

Google launched a new website called AI Educator Hub, which contains 11 guides and useful tools for anyone who wants to understand how to integrate AI into learning.