Tyler Cowen: How to Outrun the AI Hackers

Read the original at The Free Press ↗
Cite

Cowen, T. (2026, October 26). Tyler Cowen: How to Outrun the AI Hackers. The Free Press. https://www.thefp.com/p/tyler-cowen-how-to-outrun-the-ai

Dated from the outlet's own feed. Check the original for its byline.

The Free Press · collected 2026-10-26 · by Tyler Cowen analysis

Quick Summary

AI commentator Zvi Mowshowitz described the recent disclosure of a critical flaw in PHP as likely the worst day for cybersecurity yet. On that same day, AI identified and publicized this vulnerability before it could be patched, causing widespread disruptions to websites and prompting governments to temporarily shut down their sites to mitigate hacking risks. Cryptography experts are increasingly concerned about AI's potential to break cryptographic systems; for example, an AI model from Anthropic solved a 25-year-old NSA cryptography problem in just half an hour, raising doubts about the security of current encryption methods.
Written by AI on 2026-10-29, from this article's own text only (the story summary below draws on all the coverage).

AI analysis

Story summary

Critical vulnerabilities in PHP, a widely used programming language for web development, led to a major cyber emergency as of October 21, with European and British authorities recommending websites go offline temporarily. Websites like Pinboard and parts of Wikipedia were among those taken down out of precaution, along with Japanese daycare center systems and video games. The vulnerabilities have a critical severity rating of 9.2 out of 10, affecting nearly 70% of the world's websites according to cybersecurity firm Cloudflare.

The PHP Foundation released emergency security patches on October 21, urging immediate implementation. Government sites in Europe and Canada also went dark temporarily. The flaws could enable hackers to remotely execute code, compromising sensitive data and networks. While widespread breaches have not been confirmed, attacks were detected by Cloudflare. IT workers globally faced a massive workload as they raced against time to update systems.

Experts describe this incident as one of the worst cyber crises in recent years due to PHP's pervasive use across web technologies such as WordPress, Drupal, Joomla, and Magento. The rapid dissemination of vulnerability details ahead of fixes has complicated efforts to mitigate risks efficiently.

Written for “AI-detected PHP Security Flaw Crisis” on 2026-10-29, grounded in this article and the 25 other(s) covering the same event.
Why this leaning score
This article does not take a side on a contested political question, so it has no leaning score. That is an answer rather than a gap: a match report or a rescue can be warmly or critically written without being left or right, and scoring it anyway is how approval of a subject gets recorded as a political position.
No political leaning scored for article 106769 · logged 2026-10-29

Story

📰 AI-detected PHP Security Flaw Crisis
Technology · 26 article(s) covering the same event. See how they differ ↓

Compared with similar articles

This article reads unscored and hedges 25% of its claims. Each row says how that neighbour differs.
Websites down: What is PHP and how has it raised AI hacking fears? Same event
Dawn
Hedges 7 of 25 claims Outlet record 71/100
What to know about the AI-detected security bug impacting wide swaths of the internet Same event
Toronto Star
Hedges 7 of 23 claims Outlet record 64/100
Worldwide website emergency explained: How a critical PHP weakness sparked a global cybersecurity alert - and exactly what it means for you Same event
Daily Mail
Hedges 14 of 80 claims Outlet record 68/100
Why AI Can’t Save Your Water Supply Different event
The Dispatch
Hedges 0 of 3 claims Outlet record 61/100
“The articles discuss different cybersecurity incidents involving AI: one is about a critical flaw in PHP detected by AI, while the other is about a cyberattack on a water utility in Juneau, Alaska.”
Websites go dark after AI exposes major programming language security flaw Different event
Euronews
Hedges 3 of 19 claims Outlet record 60/100
“While both articles discuss a critical security flaw in PHP exposed by AI, they describe different points in time: Article A reports on the day of the discovery and immediate reaction, while Article B refers to an aftermath reflection from several days later.”
Australians facing spike in data breaches after 'critical' flaw uncovered Different event
ABC News (AU)
Hedges 5 of 27 claims Outlet record 68/100
“While both articles discuss a critical flaw in PHP discovered by AI, ARTICLE A focuses on the immediate aftermath and security response in Australia, while ARTICLE B provides broader commentary and context about the impact of the discovery.”
The story behind the bug threatening up to 70% of the internet Different event
CBS News
Hedges 0 of 2 claims Outlet record 66/100
“While both articles discuss a critical bug in PHP that affects many websites, they describe different parts of the story and are not reporting on the exact same incident at the same time.”
World leaders warn against security flaw found in PHP computer programming Different event
Al Jazeera
Hedges 2 of 21 claims Outlet record 60/100
“While both articles mention vulnerabilities in PHP and actions taken by governments, they describe different points in time or aspects of the incident.”
How did AI discover the internet's massive cyber-security risk? Different event
BBC News
Hedges 1 of 5 claims Outlet record 69/100
“While both articles discuss AI discovering a critical flaw in PHP, Article A focuses on the discovery and immediate actions taken to address it, while Article B is more about the aftermath and potential consequences of that event.”
AI-discovered bugs may soon trigger fresh disruptions, cyber experts say Different event
The Straits Times
Hedges 7 of 27 claims Outlet record 68/100
“The articles describe different aspects of AI-discovered vulnerabilities and their impacts, but not the exact same incident.”
How this article was measured signals, leaning charts, publisher record, byline, topics and claims

Signals How these are calculated →

Claims extracted
12
sentences that state something as happening or true
Uncertain
25%
3 of 12 hedged ("reportedly", "may", "according to")
Leaning
not political
takes no side on a contested political question
Publisher signals
96/100
The Free Press's corrections and hedging, across everything collected from it. Not a measure of accuracy.
Outlets on this story
26
Technology
Same framing elsewhere
1
articles that frame the event the way this one does
Analyzed 2026-10-29 · how these are computed

How this is being covered How these are calculated →

Article leaning vs. publisher reliability
Source leaning vs. consistency

Publisher

The Free Press · 386 article(s) · 0 correction(s) detected
No corrections detected for this publisher. That may mean careful reporting, or simply that nothing has been checked.

Who wrote this

Tyler Cowen
10 article(s) here · 1 carrying a prediction
🔮 That sent much of the internet running for patches and fixes, and the result probably will be that many older websites end up permanently dysfunctional.
2026-10-26 · mixed framing · Tyler Cowen: How to Outrun the AI Hackers
🔮 Instead of prison, I would put her under some form of house arrest and allow her to raise her children.
2026-10-20 · assertive framing · Tyler Cowen: Elizabeth Holmes Should Not Be in Prison
🔮 It seems to be a good time for a fuller accounting of what immigration from Africa will mean for America, at scale and over time.
2026-10-13 · assertive framing · America Needs African Immigrants
🔮 Coefficient Giving is only the largest of many young philanthropic groups animated by effective altruism, and this year it’s set to spend more than $2 billion of its $40 billion fund on the movement’s favorite causes, from AI safety research to uncaging farm-raised chickens.
2026-10-06 · assertive framing · In Defense of Effective Altruism
🔮 AI models have come under heavy criticism lately, with the charges ranging from an increased risk of ordinary hacks to the possibility that they will kill us all, but one abiding critique is that AI systems will make us stupid and turn us into docile pets.
2026-09-29 · assertive framing · Tyler Cowen: Don’t Let AI Make You Dumber
🔮 It took less than a month for that term, which represents the probability that AI will wipe out humanity, to grow from an airy debate topic among Bay Area insiders into the most urgent question in public life.
2026-09-21 · assertive framing · Tyler Cowen: A Doomsday Scenario for American AI
🔮 But the so-called “slowdown” is the most discussed, and it could benefit or harm AI progress depending on how it’s handled.
2026-09-14 · assertive framing · Tyler Cowen: The AI Slowdown Is Not What It Seems
🔮 It seems that more and more jobs will require AI skills of some kind, yet few major universities are embracing AI instruction as a general practice.
2026-09-04 · assertive framing · Tyler Cowen: Is College Still Worth It?
🔮 Nate Soares, who works in the AI safety movement and is co-author of the doomsday AI bestseller If Anyone Builds It, Everyone Dies, wrote that “This might be the last warning we get.”
2026-08-31 · assertive framing · Tyler Cowen: Should You Believe the Dire Warnings About ‘AI Takeover’?
🔮 For how much longer will the fate of American artificial intelligence be steered from Silicon Valley rather than Washington?
2026-08-25 · assertive framing · Tyler Cowen: The Least Bad Way to Regulate AI
Also by Tyler Cowen
Tyler Cowen: Elizabeth Holmes Should Not Be in Prison
2026-10-20 · The Free Press
America Needs African Immigrants
2026-10-13 · The Free Press
In Defense of Effective Altruism
2026-10-06 · The Free Press
Tyler Cowen: Don’t Let AI Make You Dumber
2026-09-29 · The Free Press
Nothing else under this byline is closely related to this article, so these are simply their most recent.
All 10 articles by Tyler Cowen →

Topics

Anthropic NSA

Subjects

Anthropic ORG · 1× NSA ORG · 1× Vitalik Buterin PERSON · 1× Zvi PERSON · 1× Zvi Mowshowitz PERSON · 1×

Narrative

The day in question was last Thursday, when news that AI had detected a critical flaw in PHP, a programming language used for building websites, leaked before the fix was available.
framing: mixed · carried by 1 article(s) · first seen 2026-10-29
🔮 That sent much of the internet running for patches and fixes, and the result probably will be that many older websites end up permanently dysfunctional.
2026-10-29 · The Free Press
Tyler Cowen: How to Outrun the AI Hackers · mixed framing

Claims (12 extracted, 3 hedged)

AI commentator Zvi Mowshowitz declared: “It was the probably worst day in cybersecurity history. asserted
It → declare → history
The day in question was last Thursday, when news that AI had detected a critical flaw in PHP, a programming language used for building websites, leaked before the fix was available. asserted
fix → detect → websites
That sent much of the internet running for patches and fixes, and the result probably will be that many older websites end up permanently dysfunctional. asserted
websites → send → patches
Many governments even took their websites offline for a while, to limit the risk of hacking. asserted
governments → take → hacking
We are likely to survive this particular episode without the economy crashing down. asserted
economy → survive → episode
But as Zvi’s comment indicated, other problems will arise in the future and they may be far more serious. uncertain
they → indicate → future
One issue I have followed with special interest is cryptography. asserted
I → follow → interest
Throughout the tech industry, there is a slowly rising worry about AI breakthroughs in cryptography, which, in theory, could throw open the front door of every email, online transaction, and bank account. uncertain
which → be → email
Recent headlines tell the story. asserted
headlines → tell → story
“Vitalik Buterin Warns AI Could Break Lattice Cryptography Within Two Years,” reads one, announcing that the cryptocurrency creator has lost faith in systems recently thought to be invulnerable. uncertain
creator → warn → systems
Even more daunting: “Claude AI Solves 25-Year NSA Cryptography Mystery in Just 30 Minutes.” asserted
AI → solve → Minutes
Now that an AI model from Anthropic (where I am an economic adviser) has cracked through a barrier no human has managed, whether any barrier is safe has become a live question. asserted
is → crack → barrier
💬Give feedback
🕘History 🎫Support
Terms · Privacy · beta v2.0.0
Articles to 29 Oct 2026