New AI Trends in 2025

According to experts, putting responsibility first will be one of the typical AI development trends in 2025.

Developing accountability is an important factor in promoting increasingly useful AI. Illustration photo: Bloomberg.

Generative AI is changing many fields. Ineducation , AI supports learning, detects plagiarism, and assists in teaching. In the medical field, AI supports image diagnosis, personalizes treatment regimens, and drug research. This tool also promotes automated content creation, supports customer care, and creates “virtual KOLs” in business and marketing.

Along with its benefits, the proliferation of generative AI also raises ethical concerns, as the tool can amplify bias and spread misinformation. AI also has the potential to compromise privacy, data security, and workforce stability.

Similar to many countries in the world , Vietnam is emerging in the application of generative AI. In order for AI to become a useful tool instead of a harmful one for humans, developing AI responsibly is one of the important tasks.

Responsible AI Trends

Dr Sam Goundar, senior lecturer in Information Technology at RMIT Vietnam, recalls some of the controversies surrounding the ethics of generative AI, from the Taylor Swift porn scandal, to the impersonation of former US President Joe Biden, to academic fraud using AI.

“More worrying are cases of chatbots inciting suicide, creating child abuse content, encouraging assassinations, biased hiring algorithms, or security vulnerabilities being exploited by AI.

These issues highlight the need for a human-centered approach to mitigate AI misuse and protect individual and societal interests,” Dr. Goundar shared.

Microsoft Copilot tool screen. Photo: Bloomberg.

According to Dr. Goundar, 2025 will see a trend toward responsible, human-centric AI. With growing concerns about bias, misinformation, and ethical risks, transparent AI (XAI) technology is prioritized, helping users understand how AI makes decisions.

Several countries such as the US, Canada, Australia, China, Japan, etc. have begun to adopt AI regulations to manage and ensure ethical AI deployment across the industry.

Another trend is Hybrid AI, a model that combines AI and humans, which may also develop in 2025.

In sustainable development, AI can help address climate change and promote green technology. In addition, AI will expand its applications in cybersecurity, helping to enhance the ability to analyze risks and respond to threats in the digital space.

“As AI adoption accelerates, ensuring responsible AI governance is essential to optimize the technology’s benefits and reduce potential risks,” Dr. Goundar emphasized.

Responsible AI Development Solutions

Vietnam is one of the countries actively innovating in the field of AI. Therefore, ensuring ethical AI development plays an important role, avoiding biased AI algorithms, privacy risks and loss of public trust.

To develop AI according to ethical principles, Dr. Goundar proposed a number of solutions such as investing in research on AI ethics, cooperating with universities to provide a legal framework for responsible AI deployment.

Next, consider integrating AI ethics into university curricula, expanding AI training programs for business leaders, educators, and policymakers.

Promoting responsibility and ethics helps AI become a useful tool for humans. Photo: Bloomberg.

Raising public awareness is also important, helping individuals and businesses understand the impact of AI on privacy. Additionally, it can encourage the application of AI for social purposes such as health , environment and education, rather than focusing solely on profit.

Besides education and awareness raising, legal frameworks also play an important role in sustaining responsible AI development.

Experts from RMIT Vietnam proposed a number of solutions, such as applying stricter regulations on AI, approaching international AI governance models, launching ethical AI certification, as well as building an AI risk classification and auditing system to assess and test high-risk systems.

“As Vietnam aims to become a regional AI powerhouse by 2030, its success will depend not only on technological advancement but also on strong regulations, ethical AI investment, and public awareness to combat potential risks,” Dr. Goundar emphasized.

Source: https://znews.vn/xu-huong-ai-moi-cua-nam-2025-post1538816.html