The fast development of artificial intelligence (AI) is creating a question that may become more important than how powerful AI can become: Who is responsible when AI goes wrong?
The question has gained renewed attention in the United States following President Donald Trump’s meeting with leading technology executives at the White House on September 29. The Major AI companies include Anthropic, Google, Meta, Nvidia, OpenAI, and xAI. They agreed to a voluntary framework involving internal controls, independent external audits, and board-level review of AI safety practices. The agreement is voluntary rather than legally enforceable, although it leaves open the possibility that some measures could eventually be incorporated into laws or regulations.
There is nothing inherently wrong with industry self-regulation. Technology companies possess technical knowledge that governments and legislators may not have. They are also directly involved in developing the systems whose risks need to be assessed. Internal testing and independent auditing can therefore be useful components of responsible AI development.
However, self-regulation should not be confused with governance. Governance involves more than establishing internal standards. It also concerns who establishes the rules, who independently monitors compliance, who has authority to intervene, and who is accountable when serious harm occurs. The current U.S. agreement provides several mechanisms for internal and external review, but questions remain about enforcement and the consequences of non-compliance.
This distinction matters because AI companies operate in a highly competitive environment. They have legitimate incentives to innovate, attract investment, develop new products, and compete internationally. At the same time, they must manage risks involving cybersecurity, privacy, misinformation, autonomous systems, and other potential harms. These objectives can sometimes pull in different directions.
The United States also views AI development through the lens of international competition, particularly with China. President Trump has repeatedly emphasized the importance of maintaining U.S. leadership in AI. We understand that competition may encourage faster innovation. However, it also raises a difficult question: Can technological leadership and effective governance be pursued simultaneously?
The answer should not be to choose between innovation and regulation as if they were opposites. A more sustainable approach would combine the strengths of several institutions. Technology companies can provide technical expertise and internal safety controls. Independent researchers and auditors can provide external scrutiny. Governments can establish enforceable minimum standards where necessary. Universities and the wider scientific community can contribute evidence, ethical analysis, and public accountability.
This is particularly important for education and research. AI is already changing how students learn, how researchers write, how academic work is evaluated, and how knowledge is disseminated. Telling researchers simply to “use AI responsibly” is insufficient. Institutions need clear standards concerning transparency, authorship, research integrity, data protection, and human oversight.
The central issue, therefore, is not whether AI should be controlled by government or left entirely to industry. The real challenge is to construct a governance system in which innovation, safety, accountability, and public interest can coexist. AI may become one of the most consequential technologies of this century. The question is no longer simply what AI can do. We must also decide who is responsible for what AI does—and who has the authority to act when it goes wrong.
Dr. Djuwari is a scholar who received an Award of Peace and Humanity from the World United Humanitarian Organization (UHO), based in the UK, in 2026. A Keynote speaker in several countries in Asia, Europe, the USA, and ASEAN. The President of the International Association of Scholarly Publishers, Editors, and Reviewers (IASPER) lives in Surabaya, Indonesia
###
