Vietnamese newsrooms budget for AI faster than they embed it

Vietnamese newsrooms appear to be moving more quickly to set aside money for artificial intelligence than to embed the technology in their core editorial systems. A study presented in Hanoi found that 60% of 25 newsroom leaders, editors and reporters surveyed said their organizations had allocated AI budgets, while 16% said AI had been integrated into their content management systems. The research, presented Oct. 22 at the international “The Digital Divide and AI Capabilities” conference, interpreted the 44-percentage-point difference as a divide between access to technology and the capacity to make it part of daily newsroom operations. It concluded that a budget or permission to use AI does not itself establish effective integration into content production. The study by Nguyen Dong Khoi and Le Thi Ngoc Suong paired the Vietnamese survey, conducted in August and September, with case analyses of NHK, Nikkei, CNA and SPH Media Trust. It identified four connected obstacles: financial resources and infrastructure; operational capability and AI-output validation; the gap between digital-transformation commitments and practice; and legal and professional-liability concerns. Risk concerns were prominent in the survey. Fifty-two percent of respondents said they had encountered AI-generated material containing fabricated or biased information, while 72% said they had become more cautious about legal risks or informational errors when applying the technology. The researchers found that sustained funding for infrastructure was a crucial condition for long-term AI deployment in the Asian case studies. They proposed a three-tiered AI-competency framework, alongside shared AI infrastructure built on reliable Vietnamese-language data sources. Other work presented at the conference pointed to wider skills constraints. A separate survey of 58 journalists and 10 in-depth interviews found greater confidence in routine digital tasks than in data analysis, digital fact-checking, information security and adapting to new technologies. Participants cited equipment, training and mentoring, time to learn and organizational support as barriers, while self-learning was their most common route to developing skills. Another study of 245 journalists found that critical thinking about AI was positively correlated with responsible use, and that digital-media literacy and algorithmic awareness were indirectly associated with responsible AI use through those critical-thinking skills. A separate interview-based assessment found that only one of five journalists worked at an organization with internal AI rules, and none of its 25 assessed skill classifications reached management or leadership level. The findings echo the cautious, piecemeal approach to AI in Asian public-interest newsrooms described in our August report. Earlier Vietnamese industry research also found that more than 60% of media agencies had applied or planned to apply AI in 2024, twice the 2023 share, but that usage was concentrated in tasks such as summarization, editing, translation and generating images or video rather than deeper research, business modelling or audience analysis. That earlier survey covered media agencies rather than the 25 participants in the new study, but it similarly found limited newsroom-wide strategic transformation. Most newsrooms spent less than 1 million VND, or about $40, a month on AI, or relied on free tools. Nguyen Quang Dong called for clear AI road maps, investment strategies, internal policies, risk management and ethical guidance, including an industry-wide code of conduct on responsible use.