Journalism begins where hype ends

,,

We’re creating alien beings. I’m willing to be selfish and choose humans"

—Geoffrey Hinton, 'Godfather of AI'

Anthropic’s Security Officer Sets 18 Months Deadline for Zero Vulnerability After PHP Crisis

When a user on X accused Clinton and his "cohorts" in AI for "creating a monster", he replied stating "yes. it was always going to be this way. we tried to warn the world."
Jason Clinton, deputy chief security officer at Anthropic against the background of Anthropic's Claude logo and illustration of a cellphone and user's hand
October 23, 2026 01:37 PM IST | Written by Vaibhav Jha

Anthropic’s deputy chief security officer (DCSO) Jason D Clinton has set a deadline of 18 months before scaling laws match up to frontier AI cyber capabilities and stop AI models from shipping security bugs, in the wake of recently discovered major flaws in open-source programming language PHP.

Taking to X, Clinton quoted an earlier tweet of his, and stated that “things are going to get a lot worse” in future and that “we need to hold on for 18 months.”

“as bad as things have gotten in the past 4 days, they are going to get a lot worse before they get better. security leaders, never let a crisis go to waste: please go to your boards and ask for help right now. not for ai, for the basics. (we need to hold on for ~18 months.),” wrote Clinton on X.

Clinton was speaking in context of the CVE-2026-109618 and CVE-2026-109619, two critical unauthenticated remote code execution (RCE) flaws, discovered by individual and OpenAI and Google embedded researchers by using frontier AI models like GPT-5.6 Sol and Gemini 4 Argon.

The two security flaws impact near 70% of the websites available on the internet thereby leading to cybersecurity agencies worldwide issuing critical advisories and recommending quick updates of users’ devices and software.

When an X user asked about the significance of 18 months, Clinton replied stating that in this time period, scaling laws (mathematical formulas that predict model’s performance on the basis of parameters, tokens and computing power) will drive cyber capabilities of frontier AI models into small, cheap models and ensure that AI stops shipping security bugs completely.

 

“vulnzero. the scaling laws will have driven cyber mythos-class capabilities into small, cheap models and all software dev everywhere will be using models for fractions of pennies on the dollar to stop shipping security bugs permanently,” wrote Clinton.

Anthropic Security Officer’s “Warn the World” Tweet Raises Eyebrows

Another tweet of Clinton talking about “warning the world” has raised eyebrows after an X user criticized Anthropic and AI companies for creating a “monster.”

“yes. it was always going to be this way. we tried to warn the world,” replied Clinton.

 

Ever since the launch of Claude Mythos, Anthropic has faced criticism over its public representatives including CEO Dario Amodei raising red flags about their own product.

Similarly, Anthropic’s demands to halt the frontier AI research and development and submission of a proposed bill to introduce AI regulatory provisions has also invited criticism.

David Sacks, Chair of the President’s Council of Advisors on Science and Technology, has referred to the past statements of Anthropic as a marketing stunt through fear-mongering.

Also Read: AI Models Help Discover Two Critical PHP Vulnerabilities Affecting 70% of the Web

Author

  • Vaibhav Jha, editor and co-founder at AI FrontPage

    Vaibhav Jha is an Editor and Co-founder of AI FrontPage. In his decade long career in journalism, Vaibhav has reported for publications including The Indian Express, Hindustan Times, and The New York Times, covering the intersection of technology, policy, and society. Outside work, he’s usually trying to persuade people to watch Anurag Kashyap films.

    LinkedIn

Related Posts

LATEST NEWS