The Library of Congress in 2010 starting archiving every public tweet published on Twitter (it even went back and grabbed tweets dating back to the platform's inception in 2006).
The Library viewed this as an opportunity to document the emergence of social media for future generations but has since come to the realization that most of what is published to Twitter is of no value to the public (it's also a lot of work).
In a recent white paper, the Library of Congress announced it will soon be changing its tweet collection strategy. From December 31, 2017, the Library will archive tweets based on the guidelines outlined in its Collections Policy Statement and associated documents. Tweets collected and archived will be "thematic and event-based," we're told. Examples cited include elections and themes of ongoing national interest.
In other words, the Library will only focus on tweets of historical significance and those with some semblance of impact.
The Library of Congress said its collection decision was partially made because the nature of Twitter has changed over time and the volume of tweets has increased dramatically since the initial agreement to archive tweets was signed. Furthermore, the Library only archives text. Considering that many modern tweets now contain images, videos and linked content, the value of text-only collecting isn't what it once was.
The Library notes that it now has the first 12 years of tweets in its possession but public access will remain embargoed until access issues can be resolved. Unfortunately, there is no projected timetable for providing public access and I wouldn't count on it happening anytime soon. After all, it's not as if the government is known for its ability to act quickly or efficiently regarding matters of this nature.