The US administration announced that it is suspending the participation of Microsoft and six other technology companies in a program that allows employees on H-1B visas to apply for green cards. The decision was presented as a measure against alleged abuses in the hiring of foreign specialists.
The list includes Adobe, Cognizant, Infosys, Tata Consultancy Services, Wipro, HCL Technologies and Capgemini. According to US Vice President J.D. Vance, companies have not advertised job openings in the US market widely enough before seeking employees from abroad.
“There is no other company in the US that has used this system more actively than Microsoft“, Vance said.According to data cited by the Associated Press, Microsoft is the fifth largest employer of H-1B visa workers, while Amazon is in first place. Vance also said that last year Microsoft laid off 6,000 US employees, received 6,300 H-1B visas and almost 3,000 green cards.
Microsoft rejected presenting the data as evidence of replacing US employees. The company pointed out that about 80 percent of the approximately 6,000 H-1B applications in the last fiscal year were for continuation or change of status of already employed employees. The new hires, who have already applied, are already legally resident in the United States and represent about 1 percent of the company's American staff.
Anthropic introduces restrictions on behavior towards Claude
In parallel with the immigration measures in the tech sector, the company Anthropic updated its policy for using the chatbot Claude. For the first time in more than a year, the rules explicitly prohibit "prolonged and wantonly offensive or cruel behavior" towards the company's models.
The restriction is aimed at extreme cases in which a user systematically behaves cruelly for no apparent purpose. Anthropic specifies that the ban does not cover ordinary irritation, criticism, arguments, dark creative plots or testing of the model. The primary enforcement mechanism remains the termination of conversations with users who engage in persistently harmful or offensive behavior.
The updated policy goes into effect on November 12, 2026. It consolidates and expands rules for fraudulent political and commercial campaigns, fake profiles, voter fraud, developing weapons software, and arming drones and other autonomous vehicles.
Anthropic also prohibits the use of Claude to track people without their consent, as well as to make decisions or recommendations about whom to investigate, arrest, or charge in criminal proceedings. The company says the changes are related to new patterns of abuse seen in the use of artificial intelligence.
Sources: apnews.com