A 2011 Burger King post about finding a restaurant near you spent part of last weekend sending people to a Wikipedia page listing Nazi concentration camps.
A McDonald's post from the same year pointed at the September 11 attacks. A 2013 Trump post about his Scottish golf courses went to a page about the Epstein files, and one about wind turbines went to something called "fortnite baby."
None of those accounts were broken into. No password was stolen, no session was taken over, nothing was posted. The Bitly link hijacking happened inside the links themselves, and every copy of them changed direction at once.
That is the part worth understanding, because this is not really a Twitter story. It is a story about a shared namespace that millions of printed QR codes, PDFs and email footers quietly depend on.
Key facts
| Field | Detail |
|---|---|
| Service | Bitly, URL shortener |
| What broke | Old bit.ly back-halves could be registered by someone else |
| First reports | 14 November 2026 |
| Bitly statement | 15 November 2026, 19:13 UTC |
| Links disabled | 15 November 2026, 19:22 UTC |
| Affected posts | Mostly 2009 to 2014 |
| Accounts breached | None |
| Cause | Not disclosed by Bitly |
- Old bit.ly back-halves were reused, so old posts now point wherever the new owner wants.
- No account was hacked. The post text is unchanged; only the link moved.
- Bitly's help page still says back-halves cannot be reused. The creation dates disagree.
- A printed QR code built on a bit.ly link is exposed to the same problem.
- Append
+to any bit.ly link to see its destination and creation date.
1. What actually happened
Bitly confirmed the problem on X at 19:13 UTC on Sunday 15 November. The company said it had "identified an issue that allowed bad actors to repurpose certain old, inactive links to redirect people to content unrelated to the original posts." It said it suspended the accounts involved and was disabling the affected links.
By 19:22 UTC, the Burger King and McDonald's links were returning Bitly's "Link Deactivated" page. CNN reported the same evening that old posts from Bill Gates, Justin Trudeau, Mike Pence, the FBI and the Department of Defense carried redirected links too.
The affected posts cluster in 2009 to 2014. Burger King told CNN it had nothing to do with the redirects, which is true, and also the entire problem.
Nothing was hacked in the usual sense. The Burger King post still says what it always said. The words are untouched. The short link inside it now obeys a stranger.
Here is the mechanism. A bit.ly link has two halves: the domain and the part after the slash, which Bitly calls the back-half. bit.ly/19BFiM is the Burger King link from 2011. The back-half is 19BFiM.
If that back-half became available, whoever registered it chose the destination. Not just for the 2011 post. For every tweet, email, PDF and printed sign that ever carried it.
🔐 Trust is the whole story here
Our OpenSSH breakdown covers a server that let a password through when it should have refused.
2. Bitly says this cannot happen
I read Bitly's help page on 16 November 2026. It was last updated on 17 August 2026, and it says this:
"Any links using our default bit.ly domain will continue to work. The back-halves used on these links cannot be re-used."
That page covers what happens when an account is deleted, including the automatic deletion Bitly applies after three years of inactivity. The promise is deliberate: your old links keep working, and nobody else can take their place.
The creation dates say otherwise.
I checked three links myself. bit.ly/19BFiM+ and bit.ly/mLWJO0+ now return HTTP 410 and a deactivated page, with no creation date at all. bit.ly/rsBqmE+ still loads, and its info page shows a CNBC video destination with a creation time of May 26 2023 14:57 UTC.
That last link sits in a post from August 2011. An independent audit published on 16 November traced it to that post, and I confirmed the 2023 timestamp on Bitly's page rather than taking it second-hand. A post from 2011 cannot contain a link that Bitly says was created in May 2023. Both cannot describe the same link. The one in that post was replaced.
Bitly has not explained how the back-halves became claimable. Anything you read that states a cause is guessing, including the plausible ones.
Why old links keep becoming available
Bitly deletes accounts that sit inactive for three years. That is a privacy policy decision, not a security one, and it makes sense on its own terms.
The trouble is what it leaves behind. Bitly's links keep working after the account is gone, so the service now holds a growing pile of live redirects whose owners no longer exist. Every abandoned account adds to it.
The 2009 to 2014 window fits that pattern. bit.ly was the default shortener of that era, and those are also the accounts most likely to have been abandoned and purged since. Bitly has not confirmed it, and the window could have other causes. It is a plausible fit, not a proven one.
One hard piece of evidence puts the problem years before this weekend. An audit of 97 links in Trump's posts from 2009 to 2013 found 48 of them re-created on 26 May 2023, between 14:55 and 15:12 UTC, one every 21 seconds for 17 minutes. All 48 point where the original posts suggest. Somebody rebuilt a batch of dead links faithfully, in one sitting, and nobody has said who.
This already happened once
In February 2025, memecoin promoters used the same trick to point old Trump and Melania links at token listings on pump.fun. Bitly said it was fixed. Seventeen months later, links created on 15 November 2026 show the same behaviour.
The pattern repeats because it pays. Claim an old link, point it at a token, let the clicks arrive, then sell into them.
3. The printed codes nobody is checking
Tweets are easy to find and easy to delete. That is why the coverage stops there.
Bitly's help pages describe QR codes differently: "Because each QR Code created in Bitly is basically a scannable form of a short link, the effect of downgrading your account is the same for QR Codes as it is for short links."
Read that again. A printed QR code is not a picture of a link. It is the link.
So the same namespace that broke a 2011 tweet is sitting on restaurant table tents, conference badges, product boxes, menus, leaflets, PDF brochures and invoice footers printed between 2010 and 2015. Nobody can edit a laminated sign. There is no dashboard for a cardboard display stand.
I have not seen a confirmed case of a hijacked printed code, and I am not going to invent one. The mechanism simply does not care whether the link is on a screen or under a layer of plastic.
If you run a business with old printed material, this is worth ten minutes of your week. Not panic. Just a look.
4. Check any bit.ly link in five seconds
Add a plus sign to the end of the link and open it in a browser. That takes you to Bitly's info page instead of the destination.
Plain link: bit.ly/rsBqmE
Info page: bit.ly/rsBqmE+
The page shows the destination URL as text and, for links that still work, the date the link was created.
Now compare dates.
- Creation date later than the post date. The link is not the original. Treat the destination as untrusted, even if it looks harmless.
- Creation date matches the post date. Probably the original link. Still check where it goes.
- HTTP 410 or "Link Deactivated". Bitly has killed the link. You cannot inspect it, and you cannot recover the date.
- No date, just a login prompt. You are outside the public info view. Open it in a private window and check again.
That last point is the quiet problem with Bitly's fix. Disabling a hijacked link also disables its history. The two Burger King and McDonald's links are now blank 410 pages.
The plus page checks the short link, not the destination. One link from the Trump sample still lists a CNBC video URL on Bitly's page, and I opened that URL myself: it now redirects to a pump.fun coin page. A clean info page is not a safety guarantee.
Save the info page before you report a link. Once Bitly disables it, the destination and the creation date are gone, and you have no record to show anyone.
5. What to do with your old links
Start with an inventory. Search your site export, your CMS and your sitemap for bit.ly, j.mp, tinyurl and rebrand.ly. Most people find fewer than they expect, and a few find one they had forgotten about in a 2012 footer.
Get the audit script
A small Python script that reads your export, finds every short link, and pulls the creation date from Bitly's info page without following the redirect. Standard library only. Nothing leaves your machine.
Show the script (142 lines)
#!/usr/bin/env python3
"""shortlink-audit: find old short links and read their creation dates.
Reads any text files (Blogger export, sitemap, HTML, CSV, Markdown), pulls out
short links, then reads the creation date from Bitly's info page without
following the redirect.
python shortlink-audit.py feed.atom --csv out.csv
python shortlink-audit.py old-posts/ --since 2015-01-01
Stdlib only. Nothing is uploaded anywhere.
"""
import argparse
import csv
import re
import sys
import urllib.error
import urllib.parse
import urllib.request
from pathlib import Path
SHORT_HOSTS = {
"bit.ly", "j.mp", "tinyurl.com", "rebrand.ly", "ow.ly", "buff.ly",
"cutt.ly", "is.gd", "v.gd", "t.co", "rb.gy", "shorturl.at",
}
BITLY_HOSTS = {"bit.ly", "j.mp"}
URL_RE = re.compile(r"https?://[^\s\"'<>\)\]}]+", re.I)
CREATED_RE = re.compile(r'item-detail--created-date"?[^>]*>\s*([^<]+?)\s*<', re.I)
DEST_RE = re.compile(r'class="item-detail--url"[^>]*href="([^"]+)"', re.I)
UA = "shortlink-audit/1.0 (+https://www.bytegeekery.com/)"
def host_of(url):
host = urllib.parse.urlsplit(url).netloc.lower()
return host[4:] if host.startswith("www.") else host
def find_links(text):
for match in URL_RE.finditer(text):
url = match.group(0).rstrip(".,;:!?)")
if host_of(url) in SHORT_HOSTS:
yield url
def read_sources(paths):
for raw in paths:
path = Path(raw)
if path.is_dir():
for child in sorted(path.rglob("*")):
if child.is_file() and child.suffix.lower() in {".html", ".htm", ".txt", ".md", ".csv", ".atom", ".xml", ".json"}:
yield child, child.read_text(encoding="utf-8", errors="replace")
elif path.is_file():
yield path, path.read_text(encoding="utf-8", errors="replace")
else:
print(f"skip, not found: {raw}", file=sys.stderr)
def probe(url):
"""Read the info page. Never follows the redirect itself."""
host = host_of(url)
if host not in BITLY_HOSTS:
return {"status": "", "created": "", "destination": "", "note": "not a Bitly link, expand with its own tool"}
info = url.rstrip("/") + "+"
req = urllib.request.Request(info, headers={"User-Agent": UA})
try:
with urllib.request.urlopen(req, timeout=20) as resp:
body = resp.read().decode("utf-8", "replace")
status = resp.status
except urllib.error.HTTPError as exc:
return {"status": str(exc.code), "created": "", "destination": "", "note": "deactivated or removed"}
except Exception as exc: # network, DNS, timeout
return {"status": "", "created": "", "destination": "", "note": type(exc).__name__}
created = CREATED_RE.search(body)
dest = DEST_RE.search(body)
return {
"status": str(status),
"created": created.group(1) if created else "",
"destination": dest.group(1) if dest else "",
"note": "" if created else "no public creation date",
}
def main(argv=None):
parser = argparse.ArgumentParser(description="List short links and their creation dates.")
parser.add_argument("sources", nargs="+", help="files or directories to scan")
parser.add_argument("--csv", help="write results here instead of stdout")
parser.add_argument("--since", help="flag links created on or after this date (YYYY-MM-DD)")
args = parser.parse_args(argv)
seen = {}
for path, text in read_sources(args.sources):
for url in find_links(text):
seen.setdefault(url, path.name)
if not seen:
print("No short links found.", file=sys.stderr)
return 0
rows = []
for url, source in seen.items():
result = probe(url)
created = result["created"]
flag = ""
if args.since and created:
try:
from datetime import datetime
made = datetime.strptime(created.replace("UTC", "").strip(), "%b %d %Y %H:%M")
if made >= datetime.strptime(args.since, "%Y-%m-%d"):
flag = "created after cutoff"
except ValueError:
pass
if result["status"] in {"410", "404"}:
flag = "deactivated"
rows.append({
"source": source,
"short_url": url,
"status": result["status"],
"created": created,
"destination": result["destination"],
"flag": flag,
"note": result["note"],
})
print(f"{url} {result['status'] or '??'} {created or '-'}", file=sys.stderr)
fields = ["source", "short_url", "status", "created", "destination", "flag", "note"]
if args.csv:
with open(args.csv, "w", newline="", encoding="utf-8") as handle:
writer = csv.DictWriter(handle, fieldnames=fields)
writer.writeheader()
writer.writerows(rows)
print(f"{len(rows)} links written to {args.csv}", file=sys.stderr)
else:
writer = csv.DictWriter(sys.stdout, fieldnames=fields)
writer.writeheader()
writer.writerows(rows)
return 0
if __name__ == "__main__":
raise SystemExit(main())
On my three test links it produced this:
| Link | Status | Created |
|---|---|---|
bit.ly/19BFiM | 410 | no date |
bit.ly/mLWJO0 | 410 | no date |
bit.ly/rsBqmE | 200 | May 26 2023 14:57 UTC |
bit.ly/thisoneisnotreal99 | 404 | no date |
Then work through the list by hand:
- Replace short links with the real destination wherever you can still edit the page. It costs nothing and it removes the dependency.
- Do not delete a post just to hide a bad link. Fix the link. Deleting sends a 404 to anyone who bookmarked it and throws away the page's history.
- Report hijacked links to Bitly and to the platform hosting the post. Keep a screenshot of the info page first.
- Use a domain you control for anything that has to stay short. Your own DNS, your own back-halves, no shared namespace.
- For print, assume the code is permanent. A printed code should never point at a shortener you do not own.
I ran the first step on this site before writing. Sixteen posts, zero third-party short links. That was luck rather than policy, and it is the reason nothing here changed direction over the weekend.
6. Frequently Asked Questions
Was Bitly hacked?
Not in the sense of someone taking over accounts. Bitly says an issue let bad actors repurpose old, inactive links. No user account was breached, and the company has not explained how the links were reclaimed.
Were the Twitter accounts hacked?
No. X was not compromised, and nobody posted from those accounts. The posts are unchanged. Only the links inside them resolve somewhere new.
Are all bit.ly links at risk?
No. Recent links with active owners are fine. The risk sits with old links whose accounts were deleted, and with anything pointing at a back-half somebody else now holds.
Do custom short domains have the same problem?
They behave differently. A branded domain keeps working while its DNS points at Bitly, and you keep control of the domain itself. That is a much better position than a shared bit.ly back-half.
Can I still use Bitly?
Yes. The dangerous part is unattended history, not new links. If you use Bitly, keep the account active and check your old links once a year.
How do I know if one of my links was hijacked?
Add a plus sign to the link and read the creation date. If it postdates the page it appears on, the link is not yours any more.
Did the affected posts get deleted?
Some did. Burger King's original became unavailable, and other accounts corrected or removed posts. That is the wrong fix, since it deletes the evidence along with the link.
7. Final Thoughts
A short link is a promise held by somebody else.
Domains have registrars, WHOIS records, renewal dates and a dispute process when ownership goes wrong. Short links have none of that. There is no public registry that tells you who owns bit.ly/19BFiM today, and no way for the original owner to reclaim it short of asking Bitly nicely.
That gap is the actual bug, and it is older than this weekend. The February 2025 incident was a warning. This one was the same warning with better material.
So the useful takeaway is not "Bitly is bad." It is that a link you cannot inspect is a link you do not control. Add the plus sign. Check the date. Replace what you can. And if you printed a QR code in 2013, go and find out where it goes now.
If you want the same lesson from the server side, our PHP2Shell breakdown is about a component everyone trusted doing something nobody expected.
Sources
- Bitly: what happens to links after account deletion
- Bitly: why accounts get deleted after three years inactive
- CNN: political and corporate X posts redirected
- Ransomnews: link-by-link audit and timeline
- The Independent: Trump tweets redirected to Epstein files
- Disruptionist: the February 2025 incident
Found this useful?
More security explainers, tested fixes, and plain-English breakdowns.
Read more on byteGeekery →