AI News Roundup – Trump administration weighs replacing Biden-era AI export controls, AI roils job market for new college graduates, Chinese universities push AI adoption by students, and more

To help you stay on top of the latest news, our AI practice group has compiled a roundup of the developments we are following.

• • Semafor reports that the Trump administration is weighing whether to replace a Biden-era rule regulating the export of AI technologies. The rule, rescinded in May, had set up a framework for export controls for sensitive AI technologies such as AI chips designed to prevent them from flowing to countries such as China or Russia. The Trump administration had originally planned to promulgate its own, harsher, rule, but shifting priorities may lead to a less hawkish approach on AI exports, especially as the Trump administration attempts to negotiate a trade deal with China. Indeed, the Trump administration’s AI Action Plan (covered in this AI Roundup last week) calls for further development of open-source and open-weight AI models, which would be available to countries such as China by their very nature. The matter of further export controls is still being debated within the White House and the Department of Commerce. 
• • The Wall Street Journal reports on AI’s impact on the job market for new college graduates in the U.S. Several firms spoken to by the WSJ are stating that entry-level positions in fields such as consulting and software engineering are increasingly being replaced by AI. For the class of 2023, labor force participation declined in the first year after graduation, which is unusual for college graduates, all signaling that entry-level positions are far more tenuous than they ever were before, even if such positions are available, leaving new graduates to compete for fewer positions against other junior employees who were laid off. However, the transition to AI in many companies has been “messy,” and new hires are expected to have AI skills as well. It remains to be seen how the job market will continue to develop if and when AI capabilities continue to grow. 
• • The MIT Technology Review reports on the rapid push by Chinese universities for their students to use AI. According to one survey, 60% of university students and faculty in China report using generative AI tools “frequently,” defined as multiple times a day or week. In contrast with Western attitudes that perceive AI as a threat, Chinese schools are treating it as “a skill to be mastered,” particularly encouraging AI literacy and staying ahead of the curve. New technology is often seen in the country as a signifier of progress and national pride, especially with the release of DeepSeek’s groundbreaking models earlier this year. As Western AI tools such as ChatGPT and Claude are unavailable in China, local versions of DeepSeek’s AI tools have often been deployed by universities for use by students and faculty. One researcher said that “the goal shouldn’t be preventing students from using AI but guiding them to harness it for effective learning and higher-order thinking.” 
• • Forbes reports on the appearance of an AI-generated model in the pages of the famed fashion magazine Vogue. A two-page spread advertisement by the fashion brand Guess contained a blonde model wearing a maxi dress, similar in style to many other ads in the magazine. However, the ad also contained a small piece of text: “Produced by Seraphinne Vallora on AI.” Seraphinne Vallora, a Paris-based creative agency, created the AI model for the ad campaign at the request of Paul Marciano, Guess’ co-founder, eliminating the need for casting, green rooms, or other components generally required to make such advertisements. Despite the disclosure of the use of AI, the ad immediately faced backlash — many readers took to social media to express their displeasure, generally with the core question of “when a perfectly posed AI-generated model replaces a living, breathing model, what do we lose?” The debate is likely to continue within the fashion industry for many months and years to come. 
• • The Economist reports on how spy and defense agencies around the world are approaching AI for intelligence work. In 2024, the U.S. government grew concerned about Chinese advances in AI, and ordered its intelligence agencies as well as the Departments of Defense and Energy to experiment with AI and “frontier” AI labs, including Anthropic, Google DeepMind, and OpenAI. This past month, under the Trump administration, the Pentagon awarded contracts to each of the three companies, as well as Elon Musk’s xAI, to experiment with “agentic” AI models that can act on behalf of users by operating computers for them. On the intelligence side of things, chatbots from Microsoft and Anthropic have been used for data analysis, often customized to suit the needs of spies. However, many agencies often desire to build wrappers around existing AI tools, which slows adoption. However, with the race to compete in AI with foreign adversaries heating up, further adoption is expected to accelerate in the coming years.