Inside the OpenAI-linked bug said to put up to 70% of the world's websites at risk

Up to 70 percent of the world's websites, an upper bound rather than a count
The scale of the claim is large, but the source gives no method or affected software behind it.
Mark

So what is the actual claim here?

Mimi

OpenAI researchers reportedly found a bug that could put up to 70 percent of the world's websites at risk. The Wall Street Journal's Robert McMillan broke it, and CBS interviewed him.

Mark

That's a huge number. Why does it matter so much?

Mimi

Because it would mean something very widely used is involved. A flaw in a common piece of software could touch a large share of the web at once.

Luke

Careful, though. We don't know what software that is. The source never names it.

Mark

So the 70 percent could be solid or shaky?

Luke

We can't tell. It's described as an upper bound, "up to," and we don't know who calculated it or how.

Mimi

And "at risk" is not "breached." Nothing here says anyone has exploited it.

Mark

Is there a fix?

Luke

Unknown. No mention of patches, vendor notifications, or timelines.

Mimi

That's the thing to watch: the technical details and any official responses.

Luke

Agreed. Until then, this is a reported claim, and the video text carries none of the evidence behind it.

  • A claim that up to seventy percent of the world's websites could be exposed has jolted attention across the technology world.
  • The alarm rests on a short video segment, leaving the flaw's mechanics and the affected software unexplained.
  • Responsibility for the discovery sits with OpenAI researchers, with the story originally broken by the Wall Street Journal.
  • Readers are waiting on technical specifics, vendor responses, and any patch that would show how serious the threat truly is.

Word has reached the public, carried by a short televised interview, that researchers at OpenAI have uncovered a software flaw said to place as much as seventy percent of the world's websites at risk. The alarm was first sounded by Robert McMillan of the Wall Street Journal, whose reporting CBS now relays, but the details that would turn a striking figure into understood fact have not yet been set down in the material available. In the long human story of trust in the machinery that connects us, such moments remind us how much rests on a few lines of code few of us will ever read, and how cautiously a large number should be held until its foundations are shown.

A short video segment from CBS News carries a striking claim: researchers at OpenAI have found a software bug that could put as much as 70 percent of the world's websites at risk. That is essentially everything the published material says, and it is thin enough to warrant care.

The segment features Robert McMillan, the Wall Street Journal reporter who first broke the story. He joins CBS to discuss what he reported, but the accompanying text does not reproduce his remarks, so the specifics remain out of view here.

What is clear is limited. The discoverers are OpenAI researchers, the scale is an upper bound rather than a firm count, and the wording speaks of websites at risk, not compromised. The source does not name the affected software, explain how the flaw works, say whether it has been exploited, or indicate whether a fix exists. It also leaves unexplained how the 70 percent figure was reached.

Those gaps matter because a claim about most of the web only gains meaning once it is tied to specific, widely used software. Until technical details, vendor statements, or patches emerge, the figure is best read as a headline claim from the Journal's reporting rather than an established measurement.

A short video segment from CBS News carries the claim: researchers at OpenAI have found a software bug that puts as much as 70 percent of the world's websites at risk. That is the whole of what the published material states, and it is worth being exact about how thin it is.

The segment features Robert McMillan, the reporter who broke the story for the Wall Street Journal. He joins CBS News to walk through the details of what he reported. The text accompanying the video does not reproduce those details, so what McMillan actually said on air is not captured in the material available here.

What can be stated plainly is limited. The discoverers are described as researchers at OpenAI. The scale is given as up to 70 percent of the world's websites, an upper bound rather than a firm count, and the wording describes websites as at risk rather than as compromised. The original reporting belongs to the Wall Street Journal, and CBS is relaying it through an interview with the reporter.

Several things the reader would naturally want are absent from the source. It does not name the software the bug affects. It does not say how the flaw works, how it was found, or whether anyone has exploited it. It gives no indication of whether a fix exists, whether vendors have been notified, or how the 70 percent figure was calculated and by whom.

Those gaps matter because the figure is large enough to demand scrutiny. A claim about a majority of the web's sites only means something once it is tied to a specific piece of widely deployed software, and the segment text does not make that connection.

What to watch next is straightforward: the technical specifics of the flaw, the identity of the affected software, and any patches or statements from vendors or from OpenAI. Until those appear, the 70 percent number is best read as a headline claim from the Journal's reporting, not an established measurement.

OpenAI researchers found a bug that puts up to 70% of the world's websites at risk (paraphrased from the CBS News segment description).
— CBS News
Envie de l'histoire complète ? Lire l'original sur CBS News ↗
Nous contacter FAQ