Your Embargo-Breaking Blog Post Isn’t Security Research. It’s Marketing Bullshit.

Almost forgot this blog existed but I am really pissed off this morning so get ready for some spice.

The hard working folks at Semgrep (my employer) got advance notice of a new rather serious vulnerability that was going to be announced within the next 24 hours. We assembled a full team, added detections and protections to our product, wrote notifications to our customers and prepped everything we would need to have an orderly and controlled response once the embargo lifted and provide maximum protection to you our users.

But then some asshats went and published their own blog about six hours early, before the vendor had released their patch!

I’m getting really tired of security companies treating vulnerability disclosure embargos like they’re optional suggestions that can be ignored whenever the marketing department smells an opportunity for a few extra clicks. Somewhere along the way, parts of the security industry decided that being first to publish a blog post about a critical vulnerability was more important than actually protecting the people running vulnerable software. We’ve turned vulnerability disclosure into a ridiculous race for headlines, complete with branded graphics, clever names, breathless LinkedIn posts, and marketing teams desperately refreshing Google to see who’s winning the SEO battle.

Here’s a newsflash: Nobody gives a shit who published the third blog post about a vulnerability. And nobody outside your marketing department cares who published the first.

But people absolutely remember who can’t be trusted.

An embargo isn’t some arbitrary restriction invented by vendors to keep researchers quiet. It’s an agreement between people trying to solve a problem before criminals can exploit it. Researchers discover vulnerabilities, maintainers develop fixes, downstream vendors prepare updates, and defenders get ready to deploy them. Ideally, everyone gets the information they need at roughly the same time, minimizing the window when attackers know about a vulnerability but users can’t protect themselves.

Does this process always work? Hell no. Vendors drag their feet, researchers get frustrated, and coordinating disclosure across dozens of organizations can resemble herding cats through a minefield.

But the system depends on trust. When a company receives privileged vulnerability information under an embargo, it’s being trusted to use that information to protect users, not to get a head start on its next marketing campaign.

Breaking that trust has consequences beyond a few hurt feelings.

Remember Heartbleed in 2014? The vulnerability became famous almost overnight, complete with its own name, logo, and website. But the disclosure also exposed serious problems with who received advance warning and who didn’t. Some organizations had time to prepare while major Linux distributors and countless other defenders were left scrambling. Google’s own account of the incident describes how the surprise public announcement accelerated its internal patching scramble.

Heartbleed wasn’t simply a case of somebody deliberately violating an agreed publication date, but it demonstrated what happens when coordinated disclosure becomes uneven and publicity moves faster than preparedness.

Then there was Spectre and Meltdown in 2018. Researchers and vendors had spent months coordinating fixes for vulnerabilities affecting modern processors. The embargo ultimately collapsed six days early as public reporting and speculation made continued secrecy increasingly impractical. Google moved up its disclosure, and vendors still preparing for the original date suddenly had to respond to a public crisis.

Again, this wasn’t necessarily some security company’s marketing team deliberately jumping the gun. It was another example of how fragile coordination becomes once information starts escaping the agreed circle.

More recently, the October 2026 PHP2Shell disclosure turned into an absolute mess. According to the PHP Foundation, Chainguard published container images containing fixes more than six hours before the coordinated release. That meant attackers could potentially reverse-engineer the patches while much of the PHP ecosystem was still waiting for its official updates. CISA officials publicly criticized the early release.

And that’s the distinction that matters. Sometimes an embargo collapses because the vulnerability is already public or actively exploited. Sometimes disclosure happens because a vendor refuses to fix a problem. And sometimes people just want to be first.

Those are not the same thing.

Anyone who knows my history knows I’ve been a supporter of full disclosure for decades. I remember when vendors routinely ignored vulnerability reports, threatened researchers with lawsuits, and treated the people finding security problems as the enemy. And those problems haven’t gone away.

Full disclosure is one of the reasons vendors eventually started taking security research seriously. Public pressure works. Embarrassment works. And sometimes dropping a zero-day on the public is exactly what needs to happen when a vendor has been sitting on a serious vulnerability, refusing to acknowledge it, or leaving users exposed without a reasonable path forward.

Disclosure should serve the people at risk, not the convenience, or the marketing departments, of the vendor. There is a massive difference between a researcher publicly disclosing a vulnerability because a vendor refuses to act and a security company violating an embargo it voluntarily agreed to because it wants to publish before its competitors. One is holding vendors accountable. The other is exploiting somebody else’s trust.

If you don’t agree with an embargo, don’t accept it. Negotiate a different timeline. Establish your own disclosure deadline. Explain publicly why you believe the existing timeline puts users at risk. But agreeing to the rules, taking advantage of early access, and then deciding the rules don’t apply to you? That’s a pretty shitty way to conduct security research.

Let’s talk about the incentives driving this nonsense. Security companies want attention. They want backlinks, search rankings, press mentions, and executives congratulating themselves on LinkedIn. I understand that. Security is a business, and businesses need customers. But when your marketing strategy starts interfering with coordinated vulnerability disclosure, you’ve crossed a line.

The researchers who discover these vulnerabilities often spend weeks or months working with maintainers, testing patches, and coordinating with dozens of organizations. Much of that work happens quietly, without applause or fancy graphics. Then some company with early access decides it needs to be the hero of the story. Suddenly, everyone else has to scramble because somebody couldn’t wait a few more hours to publish.

And let’s be clear: publishing first doesn’t make your research better. It doesn’t make your product more effective. It certainly doesn’t make your customers safer if you’ve just handed attackers information before the broader ecosystem can respond. Being first isn’t the same as being useful.

There are legitimate reasons to break an embargo. Evidence of active exploitation, a vendor deliberately concealing a dangerous vulnerability, or a disclosure agreement that has become harmful to users can all justify changing the timeline. Those decisions deserve scrutiny, but they can be defensible. Your content calendar isn’t one of those reasons.

If you have evidence that early disclosure is necessary to protect users, explain it. Coordinate where possible. Accept responsibility for the consequences. Don’t hide behind vague claims about transparency while your marketing team celebrates the traffic spike.

And if you simply can’t resist publishing before everyone else, don’t be surprised when researchers stop inviting you into coordinated disclosures. Trust is hard to earn and remarkably easy to destroy.

The security industry has spent decades fighting for the right to disclose vulnerabilities without being silenced by vendors. It would be a damn shame if we undermined that progress because a handful of companies decided that winning the race to publish was more important than protecting users.

I’ll defend a researcher’s right to drop a 0-day when circumstances justify it. I’ll also defend the right of everyone else to stop trusting companies that abuse privileged access.

Full disclosure is about accountability. Coordinated disclosure is about cooperation. Neither is supposed to be a marketing competition.