Record and publish every government request to remove a post from social media and other platforms
Steps to make it happen
- Governments and public offices sometimes ask social media companies to take down a post.
- Every such request must be made in writing, never by phone call or spoken promise.
- A public list that anyone can read shows which office asked, which post, and for what reason.
- Social media companies also publish the requests they got and whether they removed or kept each post.
- A watchdog that is independent of the government checks that the two sets of records match.
- Requests that cannot be shown right away, such as those tied to an ongoing investigation, must be made public later, after a set deadline.
- A company may refuse any request that is not in the record.
- Over time, a society where the state cannot quietly erase what people write becomes normal.
Water so far 💧 0
The proposal in detail
This proposal covers every organization with public authority: governments, ministries, police, election bodies, intelligence agencies, local governments, and others. It applies when they ask social media sites, video sites, search services, news organization websites, and similar services to delete a post, make it less visible, suspend an account, and so on. It includes not only formal orders but also informal contacts between staff members, and requests worded as "this is false information, so please consider taking action." Requests must be made in a set format that records the requesting organization, the law or rule it relies on, the type of post targeted, and the reason it was judged false. The records go into a public ledger, and the companies separately publish the requests they received and what they did in response, such as deleting the post, leaving it up, or adding a warning label. The job of comparing the two sets of records falls to an audit body or a court that is independent of the government. As exceptions, requests involving an ongoing investigation, images of child sexual abuse, or private information such as a home address do not have to be made public right away. However, there is a limit on how long they can be kept secret, and any extension needs approval from an independent body. Even when requests are published, only the key points are shown, not the full text of the post, so the false information itself is not spread again. A company has no duty to obey a request that is not in the ledger, and treating a company that refuses worse than others is banned.
BenefitsKeeps anti-disinformation work from turning into censorship
The biggest benefit is that it puts a brake on a situation where a government can erase anything just by saying it is false information. When requests are recorded and published, a government finds it harder to ask for anything except what it can later explain to citizens, the press, and the courts. If a request targets critical reporting or opposition statements, reporters and researchers who read the ledger can spot it and raise the issue. For a person whose post was deleted, knowing who asked and why gives them something to work with in filing an objection or fighting it in court. Companies also gain a legitimate reason to turn down pressure that leaves no record, which eases the position of being caught between the government and the public. In addition, when a deletion of truly dangerous disinformation is justified, people are more likely to trust it because the procedure can be seen. Considering that the suspicion "maybe the government is deleting things for its own convenience" can itself become the seed of new conspiracy theories, a transparent procedure also becomes a foundation that supports trust in the whole effort against disinformation. If each country's ledger uses the same format, it also becomes possible to compare around the world which countries make which kinds of deletion requests.
DrawbacksSlows emergency response and shows the other side your hand
The strongest point raised by those who oppose this is the risk that the response to truly dangerous disinformation will be slowed or weakened. In situations where harm spreads within hours, such as disasters, riots, or election day, the effort of preparing written documents could get in the way of a quick response. There are also worries about foreign influence operations and organized spreading of lies: publishing which tactics are known and how well they are understood would teach the other side what to guard against, and their operations would become more skillful. It also cannot be ruled out that a published ledger would itself announce the existence of the false information, drawing attention and spreading it. For both governments and companies, recording, publishing, and cross-checking is a large cost in time and money, and the burden is heavier for small companies and countries with weaker institutions. Furthermore, even if a rule requiring records is made, if those in power pressure companies through other means such as licenses, advertising, or tax audits, nothing will appear in the ledger. Thinking that publishing requests alone will protect freedom of speech is too optimistic, and there is also a concern that the appearance of "following the procedure" could make pressure that leaves no record harder to see.
Learn the background
To fight disinformation, many governments around the world now ask social media and other companies to delete posts or limit how far they spread. There is certainly information that cannot be left alone, such as life-threatening medical lies, posts giving the wrong date or time of an election, and posts that stir up violence. But the power to decide what counts as disinformation can, if misused, become a tool for silencing criticism and news reports that are inconvenient for the government. In fact, cases have been reported in countries with many kinds of systems in which posts by reporters, opposition parties, and citizens' movements were said to have been removed in the name of fighting disinformation. What makes the problem hard is that many of these requests tend to skip public procedures and are made by phone or in informal meetings. Companies are quick to comply because they fear the effect on regulation and business deals, and the person whose post is deleted cannot even learn who made the decision. Some companies voluntarily publish how many government requests they get, but the content and reasons are often unclear, and what they publish differs from company to company. To both push ahead with fighting disinformation and prevent those in power from manipulating speech, we need a system that brings the requests themselves out into the light.
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