A growing argument in AI policy circles holds that laws and rules, on their own, won't make artificial intelligence safe or trustworthy — and that governments need to think about culture and data quality too.
Writing in The Hindu, the op-ed "AI needs cultural policies, not just regulation" argues that the future of AI "will not be secured by regulation alone." Instead, the piece says, ensuring safe and trustworthy AI for everyone means balancing regulation with policies that promote high-quality data. In other words, the rules we set matter, but so does the information AI systems learn from.
That concern about how AI is built and deployed runs alongside a separate warning about who controls it and to what end. A new paper posted to arXiv, titled "From Democracies to Autocracies: How AI Systems Enable Authoritarianism by Design," argues that "AI-enabled authoritarianism is not confined to autocracies." According to the paper's abstract, the authors investigate and map the lifecycles of six AI systems deployed across different political systems, ranging from the United States to China, in an effort to provide greater transparency about how these technologies operate.
Taken together, the two sources point to a broadening conversation. One stresses the inputs — data and cultural context — that shape whether AI serves the public well. The other stresses the outputs — how AI systems can concentrate power and enable control, even in democracies, depending on how they are designed and used.
Why it matters: As policymakers race to write AI rules, these arguments suggest that regulation alone may miss the deeper questions of data quality and design intent that ultimately determine whether AI strengthens or undermines free societies.