Brian Holmes via nettime-l on Wed, 18 Nov 2026 16:03:02 +0100 (CET)


[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

Re: <nettime> Falling into the Breach


Folks, I totally agree that the infrastructure in the US is old and the
administration is vastly incompetent, making even very important systems
like air traffic control uniquely vulnerable. But the observation misses
two key things about the present situation.

The first is that there is very likely a reason for the current confusion:
it's because US officials and their private contractors are legally barred
from speaking about cybersecurity events, due to legal frameworks like
these:

-- CIRCIA (Cyber Incident Reporting for Critical Infrastructure Act):
Implemented by CISA, CIRCIA requires critical infrastructure operators
(aviation, energy, banking, telecom) to report major cyber incidents within
72 hours. Crucially, the statute classifies these reports as strictly
confidential, exempts them from Freedom of Information Act (FOIA) requests,
and prohibits agencies from sharing them publicly in any way that
identifies the victim without consent.

-- Defense Industrial Base Rules (DFARS): Contractors like L3Harris or
major aerospace firms are bound by Defense Federal Acquisition Regulation
Supplement clauses requiring rapid incident reporting to the Pentagon while
strictly prohibiting public disclosure of unclassified controlled technical
information (CTI) without DoD clearance.

The second point is that the use of AI to bring down those old, vulnerable
systems has reached a crisis level, as you can see by this FBI
cybersecurity flash that was released on Nov 13:

Threat Overview of ARTEX Autonomous Penetration Testing Framework
https://www.ic3.gov/CSA/2026/261113.pdf

Summary
The Federal Bureau of Investigation (FBI) is releasing this FLASH to
highlight the tactics, techniques, and
procedures (TTPs) and indicators of compromise (IOCs) associated with the
ARTEX autonomous
penetration testing framework. ARTEX is an open-source, large language
model (LLM)-driven agentic
platform that threat actors have weaponized to automate multiple phases of
cyber intrusions at machine
speed, from reconnaissance through data exfiltration. Since at least August
2026, a variety of malicious
cyber actors have used ARTEX to conduct campaigns targeting organizations
globally across multiple
sectors, including financial services, government, technology, defense,
semiconductor, energy,
healthcare, manufacturing, telecommunications, and legal
industries.1,2,3,4,5
ARTEX significantly reduces the time required to chain attack techniques,
with some intrusions
progressing from initial targeting to privileged access in under three
hours.² The framework enables threat
actors with minimal technical expertise to conduct complex, multi-target
operations simultaneously,
fundamentally accelerating attack velocity and expanding operational scale.
Notable campaigns include
financially motivated attacks against South Korean financial institutions
resulting in exfiltration of personal
information from approximately 67,000 individuals,6,7,8 and a supply chain
compromise affecting
ConnectWise Cloud Backup customers that resulted in attempted sign-ins to
more than 2,250
downstream tenants.2
Following public disclosure of malicious usage, the original ARTEX
developer removed the GitHub
repository in October 2026.3 However, the framework continues to
proliferate through community forks,
localized variations, archived copies, and a commercial version (ARTEX
PRO), ensuring continued
availability to threat actors.3,10 As of early November 2026, researchers
identified over 1,600 publicly
exposed ARTEX console instances, with total deployments likely
significantly higher when including
systems behind firewalls or in private networks.9
The FBI encourages organizations—particularly those operating
internet-exposed applications, self-
service portals, cloud infrastructure, and partner/supplier systems—to
implement the recommendations
in the Mitigations section to reduce the likelihood and impact of
ARTEX-enabled intrusions.

Threat
ARTEX is an autonomous, multi-agent penetration testing platform written in
Go.3,4 The framework
translates high-level tactical goals described in natural language into
actionable machine commands
executed by specialized super intelligence (SI) agents, requiring minimal
human interaction beyond initial
objective setting.2,3 ARTEX does not include its own SI model but
integrates with external LLM providers,
including frontier models (Anthropic Claude, OpenAI GPT) and open-source or
Chinese-origin models
(DeepSeek, GLM, Grok, Kimi).3,6,7,9
Analysis of attacker-controlled ARTEX instances shows Chinese models are
most prevalent, with
DeepSeek Flash family models dominating usage due to their balance of
capabilities, relatively low cost,
and lack of observable guardrails when framed for penetration testing.7,9
Originally released on GitHub by developer “Autumn-27” in July 2026, ARTEX
gained rapid adoption,
reaching an estimated 1,500–2,000 GitHub stars within 10 weeks.4 The
developer stated the tool was
intended for authorized security testing within defined scopes and made the
repository unavailable on
October 8, 2026, following reports of threat actor abuse.3,6 Despite this
action, ARTEX remains widely
accessible through independent routes and mirrors on code-sharing
platforms.3,4,9
Observed Capabilities and Potential Uses
Based on observed capabilities and ongoing threat actor adoption, ARTEX
presents the following risks:
• Accelerated Multi-Target Operations: SI agents provide automation for the
time-consuming
phases of enumeration and exploitation; small threat groups can run
concurrent intrusion
campaigns at a scale previously unavailable to them.
• Exploitation of Overlooked Attack Surfaces: ARTEX excels at identifying
and exploiting peripheral
systems receiving less security attention than flagship applications,
including partner and supplier
portals, mobile backends, employee applications, sales-support systems,
legacy services, third-
party integrations, and applications created across different business
units.1
• Supply Chain Compromise: Agents have demonstrated capability to identify
and exploit supply
chain attack vectors, including modifying iOS application distribution
infrastructure via exposed
cloud credentials.9 Autonomous agents can systematically map trust
relationships between
organizations. By identifying and compromising weaker third-party vendors,
CI/CD pipelines, or
integrated software providers, attackers can leverage ARTEX to pivot from a
peripheral supplier
directly into high-value upstream enterprise networks.
• Reduced Barrier to Entry: ARTEX lowers technical barriers for conducting
complex intrusions,
enabling threat actors with minimal expertise to execute sophisticated
campaigns.3,9
[snip]

For a broader look at the landscape of automated agentic intrusion,
including other frameworks beyond ARTEX, see this Google threat
intelligence brief from Nov 16:

https://cloud.google.com/blog/topics/threat-intelligence/agentic-threat-activity-crosses-a-significant-threshold

So, it's interesting to see in some detail how both state and corporate
actors respond to an emergency that the public isn't even allowed to know
about.... The details are out there, but you have to drill down a little. I
would be curious to hear more about all this.

On Wed, Nov 18, 2026 at 6:10 AM Felix Stalder via nettime-l <
nettime-l@lists.nettime.org> wrote:

>
> As far as I can see, there is an incredible potent mix creating the
> current vulnerabilities of infrastructures:
>
> - neglected infrastructures after a decades of neo-liberal hostility
> towards "unproductive" maintenance and general hatred of public
> investment, with Dodge and all the chainsaw guys being the meme-
> contingent of this (apparently, the router that caused the breakdown of
> the US flight control system should have been replaced in 2005).
>
> - centralization of IT infrastructure in near monopolies that makes what
> might have been routine interruptions global affairs.
>
> - geopolitical will to escalate of infrastructure attacks (such as the,
> apparently, Iranian attacks on freezer systems in Europe and the US,
> again on systems that were officially decommissioned a long time ago,
> hence unmaintained by the vendor (and, I assume, unmaintainable by the
> users, should they have wanted to, due to copyright restrictions)).
>
> - and, AI, which is both offering new capabilities but also allows
> everyone to make all kinds of effectively unverifiable claims regarding
> those capacities and possible strategies.
>
> At least for the latter, AI seems to increase the speed at which a
> vulnerability can be exploited (think of the recent PHP issues), I
> haven't seen too much evidence about it finding really new ones (rather
> than detecting known ones that have not been fixed in a particular system).
>
> That seems to be the main point of this excellent video that examined in
> detail the supposedly massive Linux kernel vulnerabilities found by AI,
> posted not too long ago here.
>
> On 22/10/2026 01:29, Louis Rawlins via nettime-l wrote:
>
> > Today, I spent some time reading through dialogues on the Project VRM
> > mailing list with Doc Searles. Somebody there shared this:
> >
> > Greg Kroah-Hartman / Kernel Recipes 2026 - Security in the LLM age
> https://
> > www.youtube.com/watch?v=NnV_cWeoo5Q
>
>
> All of this is only to reiterate Ted's point. The layers of complexity
> and nontransparency are so thick that infrastructure breakdowns become
> essentially unknowable to most of the people and institutions that
> depend on them.
>
>
> all the best. Felix
>
>
>
> On 18/11/2026 05:42, Ted Byfield via nettime-l wrote:
>
> > The people I know who reliably know about these things know nothing.
> > We saw the same thing with the wholesale destruction and abandonment
> > of US military bases across the Mideast — something seemingly
> > impossible to keep secret, and yet still very little is known about
> > it. In that case one of the US government’s levers is its power over
> > satellite imagery companies; I think similar mechanisms are probably
> > at play in these network failures. And then there’s the strange case
> > of a US submarine that recently “grounded” and, soon after, its
> > command staff was fired — Congress should have been notified of both,
> > but is only finding out now:
> >
> > https://counter.ink/blg/new/2026-11-14-uss-south-dakota-mystery
> >
> > I think we’re beginning to see just how many ways Trump & Co have
> > utterly broken the US: the ongoing corruption from the top, the
> > rampant abuses of the oligarchy, the post-DOGE skeleton-crew logic,
> > the pervasive fear among those who remain, the dissolution of channels
> > between federal agencies, the ubiquitous fear and suspicion within the
> > Pentagon, the absolute lack of trust among US allies, the witches’ brew
> > of destructive incentives reshaping how business deals with the
> government,
> > the abject failure of congressional oversight, the complete mistrust
> > across the entire field of law from to to bottom, the complete
> > destruction of meaningful bipartisan cooperation in Congress, the
> > preoccupation of the incoming Dem majority with leadership fight
> > dynamics, the collapse of cooperation between federal and state
> > governments, the pathetic subservience of the media, and just generally
> > the impossibility of reaching the one and only consensus we need at
> > this point — that the people and party in power are a stunningly
> > impossible combination of imbecilic, incompetent, delusional,
> > dishonest, and incredibly destructive. Everyone worth their salt
> > knows that the US is teetering on the edges of a dozen abysses at
> > once and is terrified that if it falls into one we’ll all fall into
> > all.
> >
> > And I didn’t even mention AI.
>
> --
> | |||||||||||||||| http://felix.openflows.com |
> | |||||||||| https://tldr.nettime.org/@festal |
> | for secure communication, please use signal |
>
> --
> # distributed via <nettime>: no commercial use without permission
> # <nettime> is a moderated mailing list for net criticism,
> # collaborative text filtering and cultural politics of the nets
> # more info: https://www.nettime.org
> # contact: nettime-l-owner@lists.nettime.org
>
-- 
# distributed via <nettime>: no commercial use without permission
# <nettime> is a moderated mailing list for net criticism,
# collaborative text filtering and cultural politics of the nets
# more info: https://www.nettime.org
# contact: nettime-l-owner@lists.nettime.org