Lawmakers have discovered a revolutionary new enforcement mechanism: threats that evaporate the moment someone says sorry. The government will legislate against child exploitation on social media immediately, or possibly never, depending entirely on whether tech firms feel like self-regulating this time around.

The arrangement is elegant in its absurdity. Parliament will draft laws, table them, debate them, and prepare to vote—unless Meta, TikTok, and YouTube send sufficiently sincere commitment letters beforehand. At which point the entire legislative process gets “reassessed.” Nobody has clarified what a reassessment looks like, or how many years of non-compliance it takes before the government remembers that laws exist.

Tech firms, naturally, are thrilled. They’ve been promising to stop children sharing nude images since roughly 2015, when the problem became impossible to ignore. Their track record suggests they will install some filters, publish a blog post about safety, and call it solved. The filters will be moderately effective at catching the obvious stuff and completely useless at everything else.

The government’s strategy relies on the assumption that this time, the promises will stick. This is the same government that believed Facebook about election interference, Instagram about mental health impacts on teenagers, and YouTube about algorithmic radicalization. Three strikes would suggest a pattern, but apparently we’re still warming up.

The legislation will probably happen eventually. Or it won’t. Depends on how convincing the next commitment letter is.