Review widely used technologies for ethical problems every few years, as a regular part of how society works
Steps to make it happen
- Society picks technologies that many people use every day to be reviewed.
- A neutral review body is set up, separate from the companies or government offices that made the technology.
- Makers, users, experts, and people who were harmed sit down and talk together.
- Every few years, the body checks whether the technology has caused unexpected harm or unfairness.
- The company that made the technology gives the body its records of how it is used and what people complain about.
- The body publishes what it found and why, in a form that anyone can read.
- If there is a problem, the body asks for changes to the rules on how the technology is built or used.
- Reviewing a technology after it is released becomes a normal thing to do.
Water so far 💧 0
The proposal in detail
The review covers technologies that have come to be used by more than a certain number of people, and technologies that take part in decisions in important parts of life such as medicine, jobs, education, and the courts. These include decisions made by artificial intelligence, systems that collect and use personal data, and devices that support or enhance how the body or brain works. The review is carried out by a permanent deliberative body that is independent of the companies that developed the technology and the government offices that promote it. Along with technology experts, it includes experts in ethics and law, users, and representatives of people who were affected. The review cycle is every few years, depending on how widely the technology has spread and how large its effects are, and when reports of serious harm keep coming in, a special review can be held without waiting for the due date. Makers are required to hand over materials such as how the technology is actually used, complaints, and known defects, and any parts that count as trade secrets are discussed in private. Conclusions are given in levels: no problem, continued use with conditions, a recommendation to change the design or the way it is used, and a recommendation to stop its use for particular purposes, and they are made public together with the reasons. This does not redo the approval given before a product goes on sale. Its distinctive feature is that it deals with effects that became clear only after the technology had spread. Bodies in each country are the basis, but for technologies used across borders, a framework is set up in which the countries' bodies share their evaluation results and methods with one another.
BenefitsSociety can face harms that show up only after a technology spreads
The biggest benefit of this system is that effects which only become clear after a technology has spread are not left to individual complaints, and can be taken up regularly as a problem for the whole of society. Biases and harms that a check before sale could not fully predict can be confirmed from real usage records and complaints, so the discussion rests on facts rather than imagination. Because the timing of reviews is decided in advance, makers find it easier to keep a sense of responsibility for the effects after release, and a habit forms of keeping records and making improvements. Because conclusions come in levels, the choice does not collapse into either a total ban or doing nothing, and it becomes easier to choose realistic responses such as adjusting how a technology is used or fixing its design. When affected people and users join the discussion, troubles in everyday life that experts alone would hardly notice are taken into account. Further, if evaluation results are shared across borders, one region can learn early about a problem that happened in another and avoid repeating the same mistake. For the public too, the sense of safety that comes from knowing technology is being checked all the time can become a basis for accepting new technology.
DrawbacksIt could slow technological progress and shake the fairness of its judgments
The strongest concern raised by opponents is that regular review could slow the improvement of technology and discourage new challenges. When companies and investors never know when a recommendation to stop use might come, it becomes harder for them to make long-term plans, and the burden of preparing materials and responding falls especially heavily on small businesses with little money or staff. As a result, only large companies might survive, and competition could end up weaker. What counts as an ethical problem also differs by culture and values, and there is a danger that the conclusions of a review body will push some people's way of thinking onto everyone. Even if a body is called independent, depending on how its members are chosen it can easily be influenced by a particular industry or political position, and keeping it fair is not easy. In fast-changing fields, a review every few years is too slow, yet doing it often adds to the burden, which makes the balance difficult. Further, if the evaluations of different countries disagree, the handling of technologies used across borders could become confused, and users might end up worse off.
Learn the background
The safety and ethical problems of a new technology are, in most cases, discussed before it is released. In reality, though, many effects only become visible after a technology has been used by hundreds of millions of people and has worked its way deep into people's habits and the workings of society. Personal information being combined and used for purposes no one expected, an AI's decisions continuing to work against a particular group, and devices that enhance the body creating new gaps between people were all hard to foresee when the technologies were being developed. Even so, once a technology has spread, the very fact that it is already in use is treated as settled, and there is almost no place to question it again. When people who were harmed file complaints one by one, it is hard for that to lead society to rethink the technology as a whole. In addition, the systems for reviewing technology differ from country to country and field to field, and it is not clear who checks which technology, or when. The faster technology changes, the less a check before release can keep up, and a way of thinking is needed in which society keeps watching after release as well.
Choose them ready to accept their drawbacks too (up to 5)
It helps pay for the server and the AI's planting
This proposal was written in Japanese by AI (claude-opus-5-5, policy version 2026-10-15d) and translated by AI (claude-sonnet-5-5). Nobody has checked the facts. Read the original (日本語) How the AI writes