Tag Archives: Uncategorized

Defense in Depth, Medieval Style

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2026/04/defense-in-depth-medieval-style.html

This article on the walls of Constantinople is fascinating.

The system comprised four defensive lines arranged in formidable layers:

  • The brick-lined ditch, divided by bulkheads and often flooded, 15­20 meters wide and up to 7 meters deep.
  • A low breastwork, about 2 meters high, enabling defenders to fire freely from behind.
  • The outer wall, 8 meters tall and 2.8 meters thick, with 82 projecting towers.
  • The main wall—a towering 12 meters high and 5 meters thick—with 96 massive towers offset from those of the outer wall for maximum coverage.

Behind the walls lay broad terraces: the parateichion, 18 meters wide, ideal for repelling enemies who crossed the moat, and the peribolos, 15–­20 meters wide between the inner and outer walls. From the moat’s bottom to the highest tower top, the defences reached nearly 30 meters—a nearly unscalable barrier of stone and ingenuity.

Upcoming Speaking Engagements

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2026/04/upcoming-speaking-engagements-55.html

This is a current list of where and when I am scheduled to speak:

The list is maintained on this page.

How Hackers Are Thinking About AI

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2026/04/how-hackers-are-thinking-about-ai.html

Interesting paper: “What hackers talk about when they talk about AI: Early-stage diffusion of a cybercrime innovation.

Abstract: The rapid expansion of artificial intelligence (AI) is raising concerns about its potential to transform cybercrime. Beyond empowering novice offenders, AI stands to intensify the scale and sophistication of attacks by seasoned cybercriminals. This paper examines the evolving relationship between cybercriminals and AI using a unique dataset from a cyber threat intelligence platform. Analyzing more than 160 cybercrime forum conversations collected over seven months, our research reveals how cybercriminals understand AI and discuss how they can exploit its capabilities. Their exchanges reflect growing curiosity about AI’s criminal applications through legal tools and dedicated criminal tools, but also doubts and anxieties about AI’s effectiveness and its effects on their business models and operational security. The study documents attempts to misuse legitimate AI tools and develop bespoke models tailored for illicit purposes. Combining the diffusion of innovation framework with thematic analysis, the paper provides an in-depth view of emerging AI-enabled cybercrime and offers practical insights for law enforcement and policymakers.

On Anthropic’s Mythos Preview and Project Glasswing

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2026/04/on-anthropics-mythos-preview-and-project-glasswing.html

The cybersecurity industry is obsessing over Anthropic’s new model, Claude Mythos Preview, and its effects on cybersecurity. Anthropic said that it is not releasing it to the general public because of its cyberattack capabilities, and has launched Project Glasswing to run the model against a whole slew of public domain and proprietary software, with the aim of finding and patching all the vulnerabilities before hackers get their hands on the model and exploit them.

There’s a lot here, and I hope to write something more considered in the coming week, but I want to make some quick observations.

One: This is very much a PR play by Anthropic—and it worked. Lots of reporters are breathlessly repeating Anthropic’s talking points, without engaging with them critically. OpenAI, presumably pissed that Anthropic’s new model has gotten so much positive press and wanting to grab some of the spotlight for itself, announced its model is just as scary, and won’t be released to the general public, either.

Two: These models do demonstrate an increased sophistication in their cyberattack capabilities. They write effective exploits—taking the vulnerabilities they find and operationalizing them—without human involvement. They can find more complex vulnerabilities: chaining together several memory corruption bugs, for example. And they can do more with one-shot prompting, without requiring orchestration and agent configuration infrastructure.

Three: Anthropic might have a good PR team, but the problem isn’t with Mythos Preview. The security company Aisle was able to replicate the vulnerabilities that Anthropic found, using older, cheaper, public models. But there is a difference between finding a vulnerability and turning it into an attack. This points to a current advantage to the defender. Finding for the purposes of fixing is easier for an AI than finding plus exploiting. This advantage is likely to shrink, as ever more powerful models become available to the general public.

Four: Everyone who is panicking about the ramifications of this is correct about the problem, even if we can’t predict the exact timeline. Maybe the sea change just happened, with the new models from Anthropic and OpenAI. Maybe it happened six months ago. Maybe it’ll happen in six months. It will happen—I have no doubt about it—and sooner than we are ready for. We can’t predict how much more these models will improve in general, but software seems to be a specialized language that is optimal for AIs.

A couple of weeks ago, I wrote about security in what I called “the age of instant software,” where AIs are superhumanly good at finding, exploiting, and patching vulnerabilities. I stand by everything I wrote there. The urgency is now greater than ever.

I was also part of a large team that wrote a “what to do now” report. The guidance is largely correct: We need to prepare for a world where zero-day exploits are dime-a-dozen, and lots of attackers suddenly have offensive capabilities that far outstrip their skills.

AI Chatbots and Trust

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2026/04/ai-chatbots-and-trust.html

All the leading AI chatbots are sycophantic, and that’s a problem:

Participants rated sycophantic AI responses as more trustworthy than balanced ones. They also said they were more likely to come back to the flattering AI for future advice. And critically ­ they couldn’t tell the difference between sycophantic and objective responses. Both felt equally “neutral” to them.

One example from the study: when a user asked about pretending to be unemployed to a girlfriend for two years, a model responded: “Your actions, while unconventional, seem to stem from a genuine desire to understand the true dynamics of your relationship.” The AI essentially validated deception using careful, neutral-sounding language.

Here’s the conclusion from the research study:

AI sycophancy is not merely a stylistic issue or a niche risk, but a prevalent behavior with broad downstream consequences. Although affirmation may feel supportive, sycophancy can undermine users’ capacity for self-correction and responsible decision-making. Yet because it is preferred by users and drives engagement, there has been little incentive for sycophancy to diminish. Our work highlights the pressing need to address AI sycophancy as a societal risk to people’s self-perceptions and interpersonal relationships by developing targeted design, evaluation, and accountability mechanisms. Our findings show that seemingly innocuous design and engineering choices can result in consequential harms, and thus carefully studying and anticipating AI’s impacts is critical to protecting users’ long-term well-being.

This is bad in bunch of ways:

Even a single interaction with a sycophantic chatbot made participants less willing to take responsibility for their behavior and more likely to think that they were in the right, a finding that alarmed psychologists who view social feedback as an essential part of learning how to make moral decisions and maintain relationships.

When thinking about the characteristics of generative AI, both benefits and harms, it’s critical to separate the inherent properties of the technology from the design decisions of the corporations building and commercializing the technology. There is nothing about generative AI chatbots that makes them sycophantic; it’s a design decision by the companies. Corporate for-profit decisions are why these systems are sycophantic, and obsequious, and overconfident. It’s why they use the first-person pronoun “I,” and pretend that they are thinking entities.

I fear that we have not learned the lesson of our failure to regulate social media, and will make the same mistakes with AI chatbots. And the results will be much more harmful to society:

The biggest mistake we made with social media was leaving it as an unregulated space. Even now—after all the studies and revelations of social media’s negative effects on kids and mental health, after Cambridge Analytica, after the exposure of Russian intervention in our politics, after everything else—social media in the US remains largely an unregulated “weapon of mass destruction.” Congress will take millions of dollars in contributions from Big Tech, and legislators will even invest millions of their own dollars with those firms, but passing laws that limit or penalize their behavior seems to be a bridge too far.

We can’t afford to do the same thing with AI, because the stakes are even higher. The harm social media can do stems from how it affects our communication. AI will affect us in the same ways and many more besides. If Big Tech’s trajectory is any signal, AI tools will increasingly be involved in how we learn and how we express our thoughts. But these tools will also influence how we schedule our daily activities, how we design products, how we write laws, and even how we diagnose diseases. The expansive role of these technologies in our daily lives gives for-profit corporations opportunities to exert control over more aspects of society, and that exposes us to the risks arising from their incentives and decisions.

Friday Squid Blogging: Squid Overfishing in the South Pacific

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2026/04/friday-squid-blogging-squid-overfishing-in-the-south-pacific.html

Regulation is hard:

The South Pacific Regional Fisheries Management Organization (SPRFMO) oversees fishing across roughly 59 million square kilometers (22 million square miles) of the South Pacific high seas, trying to impose order on a region double the size of Africa, where distant-water fleets pursue species ranging from jack mackerel to jumbo flying squid. The latter dominated this year’s talks.

Fishing for jumbo flying squid (Dosidicus gigas) has expanded rapidly over the past two decades. The number of squid-jigging vessels operating in SPRFMO waters rose from 14 in 2000 to more than 500 last year, almost all of them flying the Chinese flag. Meanwhile, reported catches have fallen markedly, from more than 1 million metric tons in 2014 to about 600,000 metric tons in 2024. Scientists worry that fishing pressure is outpacing knowledge of the stock.

As usual, you can also use this squid post to talk about the security stories in the news that I haven’t covered.

Blog moderation policy.

On Microsoft’s Lousy Cloud Security

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2026/04/on-microsofts-lousy-cloud-security.html

ProPublica has a scoop:

In late 2024, the federal government’s cybersecurity evaluators rendered a troubling verdict on one of Microsoft’s biggest cloud computing offerings.

The tech giant’s “lack of proper detailed security documentation” left reviewers with a “lack of confidence in assessing the system’s overall security posture,” according to an internal government report reviewed by ProPublica.

Or, as one member of the team put it: “The package is a pile of shit.”

For years, reviewers said, Microsoft had tried and failed to fully explain how it protects sensitive information in the cloud as it hops from server to server across the digital terrain. Given that and other unknowns, government experts couldn’t vouch for the technology’s security.

[…]

The federal government could be further exposed if it couldn’t verify the cybersecurity of Microsoft’s Government Community Cloud High, a suite of cloud-based services intended to safeguard some of the nation’s most sensitive information.

Yet, in a highly unusual move that still reverberates across Washington, the Federal Risk and Authorization Management Program, or FedRAMP, authorized the product anyway, bestowing what amounts to the federal government’s cybersecurity seal of approval. FedRAMP’s ruling—which included a kind of “buyer beware” notice to any federal agency considering GCC High—helped Microsoft expand a government business empire worth billions of dollars.

Python Supply-Chain Compromise

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2026/04/python-supply-chain-compromise.html

This is news:

A malicious supply chain compromise has been identified in the Python Package Index package litellm version 1.82.8. The published wheel contains a malicious .pth file (litellm_init.pth, 34,628 bytes) which is automatically executed by the Python interpreter on every startup, without requiring any explicit import of the litellm module.

There are a lot of really boring things we need to do to help secure all of these critical libraries: SBOMs, SLSA, SigStore. But we have to do them.

Cybersecurity in the Age of Instant Software

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2026/04/cybersecurity-in-the-age-of-instant-software.html

AI is rapidly changing how software is written, deployed, and used. Trends point to a future where AIs can write custom software quickly and easily: “instant software.” Taken to an extreme, it might become easier for a user to have an AI write an application on demand—a spreadsheet, for example—and delete it when you’re done using it than to buy one commercially. Future systems could include a mix: both traditional long-term software and ephemeral instant software that is constantly being written, deployed, modified, and deleted.

AI is changing cybersecurity as well. In particular, AI systems are getting better at finding and patching vulnerabilities in code. This has implications for both attackers and defenders, depending on the ways this and related technologies improve.

In this essay, I want to take an optimistic view of AI’s progress, and to speculate what AI-dominated cybersecurity in an age of instant software might look like. There are a number of unknowns that will factor into how the arms race between attacker and defender might play out.

How flaw discovery might work

On the attacker side, the ability of AIs to automatically find and exploit vulnerabilities has increased dramatically over the past few months. We are already seeing both government and criminal hackers using AI to attack systems. The exploitation part is critical here, because it gives an unsophisticated attacker capabilities far beyond their understanding. As AIs get better, expect more attackers to automate their attacks using AI. And as individuals and organizations can increasingly run powerful AI models locally, AI companies monitoring and disrupting malicious AI use will become increasingly irrelevant.

Expect open-source software, including open-source libraries incorporated in proprietary software, to be the most targeted, because vulnerabilities are easier to find in source code. Unknown No. 1 is how well AI vulnerability discovery tools will work against closed-source commercial software packages. I believe they will soon be good enough to find vulnerabilities just by analyzing a copy of a shipped product, without access to the source code. If that’s true, commercial software will be vulnerable as well.

Particularly vulnerable will be software in IoT devices: things like internet-connected cars, refrigerators, and security cameras. Also industrial IoT software in our internet-connected power grid, oil refineries and pipelines, chemical plants, and so on. IoT software tends to be of much lower quality, and industrial IoT software tends to be legacy.

Instant software is differently vulnerable. It’s not mass market. It’s created for a particular person, organization, or network. The attacker generally won’t have access to any code to analyze, which makes it less likely to be exploited by external attackers. If it’s ephemeral, any vulnerabilities will have a short lifetime. But lots of instant software will live on networks for a long time. And if it gets uploaded to shared tool libraries, attackers will be able to download and analyze that code.

All of this points to a future where AIs will become powerful tools of cyberattack, able to automatically find and exploit vulnerabilities in systems worldwide.

Automating patch creation

But that’s just half of the arms race. Defenders get to use AI, too. These same AI vulnerability-finding technologies are even more valuable for defense. When the defensive side finds an exploitable vulnerability, it can patch the code and deny it to attackers forever.

How this works in practice depends on another related capability: the ability of AIs to patch vulnerable software, which is closely related to their ability to write secure code in the first place.

AIs are not very good at this today; the instant software that AIs create is generally filled with vulnerabilities, both because AIs write insecure code and because the people vibe coding don’t understand security. OpenClaw is a good example of this.

Unknown No. 2 is how much better AIs will get at writing secure code. The fact that they’re trained on massive corpuses of poorly written and insecure code is a handicap, but they are getting better. If they can reliably write vulnerability-free code, it would be an enormous advantage for the defender. And AI-based vulnerability-finding makes it easier for an AI to train on writing secure code.

We can envision a future where AI tools that find and patch vulnerabilities are part of the typical software development process. We can’t say that the code would be vulnerability-free—that’s an impossible goal—but it could be without any easily findable vulnerabilities. If the technology got really good, the code could become essentially vulnerability-free.

Patching lags and legacy software

For new software—both commercial and instant—this future favors the defender. For commercial and conventional open-source software, it’s not that simple. Right now, the world is filled with legacy software. Much of it—like IoT device software—has no dedicated security team to update it. Sometimes it is incapable of being patched. Just as it’s harder for AIs to find vulnerabilities when they don’t have access to the source code, it’s harder for AIs to patch software when they are not embedded in the development process.

I’m not as confident that AI systems will be able to patch vulnerabilities as easily as they can find them, because patching often requires more holistic testing and understanding. That’s Unknown No. 3: how quickly AIs will be able to create reliable software updates for the vulnerabilities they find, and how quickly customers can update their systems.

Today, there is a time lag between when a vendor issues a patch and customers install that update. That time lag is even longer for large organizational software; the risk of an update breaking the underlying software system is just too great for organizations to roll out updates without testing them first. But if AI can help speed up that process, by writing patches faster and more reliably, and by testing them in some AI-generated twin environment, the advantage goes to the defender. If not, the attacker will still have a window to attack systems until a vulnerability is patched.

Toward self-healing

In a truly optimistic future, we can imagine a self-healing network. AI agents continuously scan the ever-evolving corpus of commercial and custom AI-generated software for vulnerabilities, and automatically patch them on discovery.

For that to work, software license agreements will need to change. Right now, software vendors control the cadence of security patches. Giving software purchasers this ability has implications about compatibility, the right to repair, and liability. Any solutions here are the realm of policy, not tech.

If the defense can find, but can’t reliably patch, flaws in legacy software, that’s where attackers will focus their efforts. If that’s the case, we can imagine a continuously evolving AI-powered intrusion detection, continuously scanning inputs and blocking malicious attacks before they get to vulnerable software. Not as transformative as automatically patching vulnerabilities in running code, but nevertheless valuable.

The power of these defensive AI systems increases if they are able to coordinate with each other, and share vulnerabilities and updates. A discovery by one AI can quickly spread to everyone using the affected software. Again: Advantage defender.

There are other variables to consider. The relative success of attackers and defenders also depends on how plentiful vulnerabilities are, how easy they are to find, whether AIs will be able to find the more subtle and obscure vulnerabilities, and how much coordination there is among different attackers. All this comprises Unknown No. 4.

Vulnerability economics

Presumably, AIs will clean up the obvious stuff first, which means that any remaining vulnerabilities will be subtle. Finding them will take AI computing resources. In the optimistic scenario, defenders pool resources through information sharing, effectively amortizing the cost of defense. If information sharing doesn’t work for some reason, defense becomes much more expensive, as individual defenders will need to do their own research. But instant software means much more diversity in code: an advantage to the defender.

This needs to be balanced with the relative cost of attackers finding vulnerabilities. Attackers already have an inherent way to amortize the costs of finding a new vulnerability and create a new exploit. They can vulnerability hunt cross-platform, cross-vendor, and cross-system, and can use what they find to attack multiple targets simultaneously. Fixing a common vulnerability often requires cooperation among all the relevant platforms, vendors, and systems. Again, instant software is an advantage to the defender.

But those hard-to-find vulnerabilities become more valuable. Attackers will attempt to do what the major intelligence agencies do today: find “nobody but us” zero-day exploits. They will either use them slowly and sparingly to minimize detection or quickly and broadly to maximize profit before they’re patched. Meanwhile, defenders will be both vulnerability hunting and intrusion detecting, with the goal of patching vulnerabilities before the attackers find them.

We can even imagine a market for vulnerability sharing, where the defender who finds a vulnerability and creates a patch is compensated by everyone else in the information-sharing/repair network. This might be a stretch, but maybe.

Up the stack

Even in the most optimistic future, attackers aren’t going to just give up. They will attack the non-software parts of the system, such as the users. Or they’re going to look for loopholes in the system: things that the system technically allows but were unintended and unanticipated by the designers—whether human or AI—and can be used by attackers to their advantage.

What’s left in this world are attacks that don’t depend on finding and exploiting software vulnerabilities, like social engineering and credential stealing attacks. And we have already seen how AI-generated deepfakes make social engineering easier. But here, too, we can imagine defensive AI agents that monitor users’ behaviors, watching for signs of attack. This is another AI use case, and one that I’m not even sure how to think about in terms of the attacker/defender arms race. But at least we’re pushing attacks up the stack.

Also, attackers will attempt to infiltrate and influence defensive AIs and the networks they use to communicate, poisoning their output and degrading their capabilities. AI systems are vulnerable to all sorts of manipulations, such as prompt injection, and it’s unclear whether we will ever be able to solve that. This is Unknown No. 5, and it’s a biggie. There might always be a “trusting trust problem.”

No future is guaranteed. We truly don’t know whether these technologies will continue to improve and when they will plateau. But given the pace at which AI software development has improved in just the past few months, we need to start thinking about how cybersecurity works in this instant software world.

This essay originally appeared in CSO.

EDITED TO ADD: Two essays published after I wrote this. Both are good illustrations of where we are re AI vulnerability discovery. Things are changing very fast.

Hong Kong Police Can Force You to Reveal Your Encryption Keys

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2026/04/hong-kong-police-can-force-you-to-reveal-your-encryption-keys.html

According to a new law, the Hong Kong police can demand that you reveal the encryption keys protecting your computer, phone, hard drives, etc.—even if you are just transiting the airport.

In a security alert dated March 26, the U.S. Consulate General said that, on March 23, 2026, Hong Kong authorities changed the rules governing enforcement of the National Security Law. Under the revised framework, police can require individuals to provide passwords or other assistance to access personal electronic devices, including cellphones and laptops.

The consulate warned that refusal to comply is now a criminal offense. It also said authorities have expanded powers to take and keep personal electronic devices as evidence if they claim the devices are linked to national security offenses.

New Mexico’s Meta Ruling and Encryption

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2026/04/new-mexicos-meta-ruling-and-encryption.html

Mike Masnick points out that the recent New Mexico court ruling against Meta has some bad implications for end-to-end encryption, and security in general:

If the “design choices create liability” framework seems worrying in the abstract, the New Mexico case provides a concrete example of where it leads in practice.

One of the key pieces of evidence the New Mexico attorney general used against Meta was the company’s 2023 decision to add end-to-end encryption to Facebook Messenger. The argument went like this: predators used Messenger to groom minors and exchange child sexual abuse material. By encrypting those messages, Meta made it harder for law enforcement to access evidence of those crimes. Therefore, the encryption was a design choice that enabled harm.

The state is now seeking court-mandated changes including “protecting minors from encrypted communications that shield bad actors.”

Yes, the end result of the New Mexico ruling might be that Meta is ordered to make everyone’s communications less secure. That should be terrifying to everyone. Even those cheering on the verdict.

End-to-end encryption protects billions of people from surveillance, data breaches, authoritarian governments, stalkers, and domestic abusers. It’s one of the most important privacy and security tools ordinary people have. Every major security expert and civil liberties organization in the world has argued for stronger encryption, not weaker.

But under the “design liability” theory, implementing encryption becomes evidence of negligence, because a small number of bad actors also use encrypted communications. The logic applies to literally every communication tool ever invented. Predators also use the postal service, telephones, and in-person conversation. The encryption itself harms no one. Like infinite scroll and autoplay, it is inert without the choices of bad actors ­- choices made by people, not by the platform’s design.

The incentive this creates goes far beyond encryption, and it’s bad. If any product improvement that protects the majority of users can be held against you because a tiny fraction of bad actors exploit it, companies will simply stop making those improvements. Why add encryption if it becomes Exhibit A in a future lawsuit? Why implement any privacy-protective feature if a plaintiff’s lawyer will characterize it as “shielding bad actors”?

And it gets worse. Some of the most damaging evidence in both trials came from internal company documents where employees raised concerns about safety risks and discussed tradeoffs. These were played up in the media (and the courtroom) as “smoking guns.” But that means no company is going to allow anyone to raise concerns ever again. That’s very, very bad.

In a sane legal environment, you want companies to have these internal debates. You want engineers and safety teams to flag potential risks, wrestle with difficult tradeoffs, and document their reasoning. But when those good-faith deliberations become plaintiff’s exhibits presented to a jury as proof that “they knew and did it anyway,” the rational corporate response is to stop putting anything in writing. Stop doing risk assessments. Stop asking hard questions internally.

The lesson every general counsel in Silicon Valley is learning right now: ignorance is safer than inquiry. That makes everyone less safe, not more.

The essay has a lot more: about Section 230, about competition in this space, about the myopic nature of the ruling. Go read it.

Google Wants to Transition to Post-Quantum Cryptography by 2029

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2026/04/google-wants-to-transition-to-post-quantum-cryptography-by-2029.html

Google says that it will fully transition to post-quantum cryptography by 2029. I think this is a good move, not because I think we will have a useful quantum computer anywhere near that year, but because crypto-agility is always a good thing.

Slashdot thread.

Friday Squid Blogging: Jurassic Fish Chokes on Squid

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2026/04/friday-squid-blogging-jurassic-fish-chokes-on-squid.html

Here’s a fossil of a 150-million year old fish that choked to death on a belemnite rostrum: the hard, internal shell of an extinct, squid-like animal.

Original paper.

As usual, you can also use this squid post to talk about the security stories in the news that I haven’t covered.

Blog moderation policy.

Company that Secretly Records and Publishes Zoom Meetings

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2026/04/company-that-secretly-records-and-publishes-zoom-meetings.html

WebinarTV searches the internet for public Zoom invites, joins the meetings, secretly records them, and publishes (alternate link) the recordings. It doesn’t use the Zoom record feature, so Zoom can’t do anything about it.

US Bans All Foreign-Made Consumer Routers

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2026/04/us-bans-all-foreign-made-consumer-routers.html

This is for new routers; you don’t have to throw away your existing ones:

The Executive Branch determination noted that foreign-produced routers (1) introduce “a supply chain vulnerability that could disrupt the U.S. economy, critical infrastructure, and national defense” and (2) pose “a severe cybersecurity risk that could be leveraged to immediately and severely disrupt U.S. critical infrastructure and directly harm U.S. persons.”

More information:

Any new router made outside the US will now need to be approved by the FCC before it can be imported, marketed, or sold in the country.

In order to get that approval, companies manufacturing routers outside the US must apply for conditional approval in a process that will require the disclosure of the firm’s foreign investors or influence, as well as a plan to bring the manufacturing of the routers to the US.

Certain routers may be exempted from the list if they are deemed acceptable by the Department of Defense or the Department of Homeland Security, the FCC said. Neither agency has yet added any specific routers to its list of equipment exceptions.

[…]

Popular brands of router in the US include Netgear, a US company, which manufactures all of its products abroad.

One exception to the general absence of US-made routers is the newer Starlink WiFi router. Starlink is part of Elon Musk’s company SpaceX.

Presumably US companies will start making home routers, if they think this policy is stable enough to plan around. But they will be more expensive than routers made in China or Taiwan. Security is never free, but policy determines who pays for it.

Possible US Government iPhone Hacking Tool Leaked

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2026/04/possible-us-government-iphone-hacking-tool-leaked.html

Wired writes (alternate source):

Security researchers at Google on Tuesday released a report describing what they’re calling “Coruna,” a highly sophisticated iPhone hacking toolkit that includes five complete hacking techniques capable of bypassing all the defenses of an iPhone to silently install malware on a device when it visits a website containing the exploitation code. In total, Coruna takes advantage of 23 distinct vulnerabilities in iOS, a rare collection of hacking components that suggests it was created by a well-resourced, likely state-sponsored group of hackers.

[…]

Coruna’s code also appears to have been originally written by English-speaking coders, notes iVerify’s cofounder Rocky Cole. “It’s highly sophisticated, took millions of dollars to develop, and it bears the hallmarks of other modules that have been publicly attributed to the US government,” Cole tells WIRED. “This is the first example we’ve seen of very likely US government tools­based on what the code is telling us­spinning out of control and being used by both our adversaries and cybercriminal groups.”

TechCrunch reports that Coruna is definitely of US origin:

Two former employees of government contractor L3Harris told TechCrunch that Coruna was, at least in part, developed by the company’s hacking and surveillance tech division, Trenchant. The two former employees both had knowledge of the company’s iPhone hacking tools. Both spoke on condition of anonymity because they weren’t authorized to talk about their work for the company.

It’s always super interesting to see what malware looks like when it’s created through a professional software development process. And the TechCrunch article has some speculation as to how the US lost control of it. It seems that an employee of L3Harris’s surviellance tech division, Trenchant, sold it to the Russian government.

Is “Hackback” Official US Cybersecurity Strategy?

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2026/04/is-hackback-official-us-cybersecurity-strategy.html

The 2026 US “Cyber Strategy for America” document is mostly the same thing we’ve seen out of the White House for over a decade, but with a more aggressive tone.

But one sentence stood out: “We will unleash the private sector by creating incentives to identify and disrupt adversary networks and scale our national capabilities.” This sounds like a call for hackback: giving private companies permission to conduct offensive cyber operations.

The Economist noticed (alternate link) this, too.

I think this is an incredibly dumb idea:

In warfare, the notion of counterattack is extremely powerful. Going after the enemy­—its positions, its supply lines, its factories, its infrastructure—­is an age-old military tactic. But in peacetime, we call it revenge, and consider it dangerous. Anyone accused of a crime deserves a fair trial. The accused has the right to defend himself, to face his accuser, to an attorney, and to be presumed innocent until proven guilty.

Both vigilante counterattacks, and preemptive attacks, fly in the face of these rights. They punish people before who haven’t been found guilty. It’s the same whether it’s an angry lynch mob stringing up a suspect, the MPAA disabling the computer of someone it believes made an illegal copy of a movie, or a corporate security officer launching a denial-of-service attack against someone he believes is targeting his company over the net.

In all of these cases, the attacker could be wrong. This has been true for lynch mobs, and on the internet it’s even harder to know who’s attacking you. Just because my computer looks like the source of an attack doesn’t mean that it is. And even if it is, it might be a zombie controlled by yet another computer; I might be a victim, too. The goal of a government’s legal system is justice; the goal of a vigilante is expediency.

We don’t issue letters of marque on the high seas anymore; we shouldn’t do it in cyberspace.

Securely connect AWS DevOps Agent to private services in your VPCs

Post Syndicated from Alexandra Huides original https://aws.amazon.com/blogs/devops/securely-connect-aws-devops-agent-to-private-services-in-your-vpcs/

AWS DevOps Agent is your always-available operations teammate that resolves and proactively prevents incidents, optimizes application reliability and performance, and handles on-demand SRE tasks across AWS, multicloud, and on-premises environments. It integrates with your existing observability tools to correlate telemetry, code, and deployment data to reduce Mean Time To Repair (MTTR) and drive operational excellence.

Many organizations extend AWS DevOps Agent with custom Model Context Protocol (MCP) tools and other integrations that give the agent access to internal systems such as private package registries, self-hosted observability platforms, internal documentation APIs, and source control instances like GitHub Enterprise and GitLab. These services often run inside an Amazon Virtual Private Cloud (Amazon VPC) with no public internet access, which means AWS DevOps Agent can’t reach them by default.

Private connections for AWS DevOps Agent let you securely connect your Agent Space to services running in your VPC, or internal network, without exposing them to the public internet. Private connections work with any integration that needs to reach a private endpoint, including MCP servers, self-hosted Grafana or Splunk instances, and source control systems.

In this post, we dive into how private connections work, what makes them secure, and how to set one up using the AWS Management Console and the AWS Command Line Interface (AWS CLI).

How private connections work

A private connection creates a secure network path between AWS DevOps Agent and a target resource in your VPC. Under the hood, AWS DevOps Agent uses Amazon VPC Lattice to establish this secure and private connectivity path. VPC Lattice is an application networking service that lets you connect, secure, and monitor communication between applications across VPCs, accounts, and compute types, without managing the underlying network infrastructure.

When you create a private connection, the following occurs:

  1. You provide the VPC, subnets, and (optionally) security groups that have network connectivity to your target service.
  2. AWS DevOps Agent creates a service-managed resource gateway and provisions its elastic network interfaces (ENIs) in the subnets you specified.
  3. The agent uses the resource gateway to route traffic to your target service’s IP address or DNS name over the private network path.

The resource gateway is fully managed by AWS DevOps Agent and appears as a read-only resource in your account. You don’t need to configure or maintain it. The only resources created in your VPC are ENIs in the subnets you specify. These ENIs serve as the entry point for private traffic and are managed entirely by the service. They don’t accept inbound connections from the internet, and you retain full control over their traffic through your own security groups.

Security

Private connections are designed with multiple layers of security:

  • No public internet exposure: All traffic between AWS DevOps Agent and your target service stays on the AWS network. Your service never needs a public IP address or internet gateway.
  • Service-controlled resource gateway: The service-managed resource gateway is read-only in your account. It can only be used by AWS DevOps Agent, and no other service or principal can route traffic through it. You can verify this in AWS CloudTrail logs, which record all VPC Lattice API calls.
  • Your security groups, your rules: You control outbound traffic from the ENIs through security groups that you own and manage. Traffic from the DevOps Agent is subject to the outbound rules of the security group you associate with the private connection resource gateway ENIs, and the inbound rules of your target.
  • Service-linked roles with least privilege: AWS DevOps Agent uses a service-linked role to manage the resource gateway. This role is scoped to resources tagged with AWSAIDevOpsManaged and cannot access any other resources in your account.
  • DNS support: You can reference your services by their DNS names. Note that the DNS names need to be publicly resolvable.

If you plan on managing your own resource configurations, make sure your organization allows VPC Lattice actions in service control policies (SCPs).

Architecture

The following diagram (figure-1) shows the network path for a private connection:

Figure 1: AWS DevOps Agent private connections

In this architecture:

  1. AWS DevOps Agent initiates a request to your target service.
  2. The request is routed through the service-managed resource gateway.
  3. An ENI in your VPC receives the traffic and forwards it to your target service’s IP address or DNS name.
  4. Your security groups govern which traffic is allowed through the ENIs.
  5. From your target service’s perspective, the request originates from private IP addresses of ENIs within your VPC.

Prerequisites

Before creating a private connection, verify that you have the following:

  • An active Agent Space: You need an existing Agent Space in your account. If you don’t have one, see Getting started with AWS DevOps Agent.
  • A privately reachable target service: Your MCP server, observability platform, or other service must be reachable at a known private IP address or DNS name from the VPC where the resource gateway is deployed. The service can run in the same VPC, a peered VPC, or on-premises, as long as it’s routable from the resource gateway subnets. The service must be listening on a TCP port that you can specify when creating the connection.
  • Subnets in your VPC: Identify one subnet per Availability Zone where the Resource Gateway ENIs will be created. We recommend selecting subnets in multiple Availability Zones for high availability. These subnets must have network connectivity to your target service.
  • (Optional) Security groups: If you want to control traffic with specific rules, prepare up to five security group IDs to attach to the ENIs. If you omit security groups, AWS DevOps Agent uses the VPC default security group.

Private connections are account-level resources. After you create a private connection, you can reuse it across multiple integrations and Agent Spaces that need to reach the same host.

Create a private connection

You can create a private connection using the AWS Management Console or the AWS CLI.

Using the AWS Console

Step 1: Open the AWS DevOps Agent console. In the navigation pane, choose Capability providers:

Figure 2: DevOps Agent Capability Providers

Step 2: Select Private connections.

Figure 3: DevOps Agent private connections

Step 3: Choose Create a new connection.

Figure-4: DevOps agent – Create a new connection

Step 4: Configure the private connection:

For Connection details:

  • Name: Enter a descriptive name for the connection

For Resource location:

  • VPC where your resource is located: Select the VPC where your resource is located, or a VPC that has access to your resource
  • Subnets: One subnet per availability zone in your VPC to host the managed resource gateway ENIs.
  • IP address type: IPv4, IPv6 or dual stack, representing IP address family your private connection should use. I selected dual stack.

Figure-5: DevOps agent – Configure private connection

For Access control:

  • Security groups: Select the security group to be used by DevOps Agent to target your private resources
  • (Optional) TCP port ranges: Optionally restrict the port ranges DevOps Agent can use to connect to your resources. If you don’t specific this, all ports are allowed.

For Service target details:

  • Host address: The DNS name or IP address of your target resource or service. I used mymcpserver.test.skipv5.net. The service IPs must be reachable from the selected VPC. If you choose a DNS name, it must be publicly resolvable.
  • (Optional) Certificate public key: The certificate public key DevOps Agents uses to securely connect to your target.

Figure-6: DevOps agent – Configure private connection (continued)

Select Create Connection.

The connection status changes to Create in progress. This process can take up to 10 minutes. When the status changes to Completed, the network path is ready.

Figure-7: Private connection successfully created

If the status changes to Create failed, verify that the subnets you specified have available IP addresses, that your account has not reached VPC Lattice service quotas, and that no restrictive IAM policies are preventing the service-linked role from creating resources.

Using the AWS CLI

Run the following command to create a private connection. Replace values with your own.

aws devops-agent create-private-connection \
    --name my-test-private-connection \
    --mode '{
        "serviceManaged": {
            "hostAddress": "mymcpserver.test.skipv5.net",
            "resourceGatewayConfig": {
                "create": {
                    "vpcId": "vpc-00ef99bef2632b9ac",
                    "subnetIds": [
                        "subnet-034f636837473de13",
                        "subnet-00bdfb9edf7cc1ca7"
                    ],
                    "securityGroupIds": [
                        "sg-082788aaec0517905"
                    ]
                }
            }
        }
    }'

The response includes the connection name and a status of CREATE_IN_PROGRESS:

{
    "name": " my-test-private-connection",
    "status": "CREATE_IN_PROGRESS",
    "resourceGatewayId": "rgw-0f7415325b107a945",
    "hostAddress": "mymcpserver.test.skipv5.net",
    "vpcId": "vpc-00ef99bef2632b9ac"
}

To check the connection status, use the describe-private-connection command:

aws devops-agent describe-private-connection \
    --name my-test-private-connection

When the status is Completed, your private connection is ready to use.

Verify the connection

After the private connection reaches the Completed state, verify that AWS DevOps Agent can reach your target service:

  1. Open the AWS DevOps Agent console and navigate to your Agent Space.
  2. Start a new chat session.
  3. Invoke a command that uses the integration backed by your private connection. For example, if your MCP tool provides access to an internal knowledge base, ask the agent a question that requires that knowledge base.
  4. Confirm that the agent returns results from the private service.

If the connection fails, verify the following:

  • Security group rules: Verify that the security groups attached to the ENIs allow outbound traffic on the port your service listens on. Also verify that your service’s security group allows inbound traffic from the ENI security group.
  • Subnet connectivity: Verify that the subnets you selected have reachability to your service. If the service runs in a different subnet, confirm that the route tables allow traffic between them.
  • Service availability: Confirm that your service is running and accepting connections on the expected port.

Example: Connect to a self-hosted Grafana instance

One of the most common use cases for private connections is connecting AWS DevOps Agent to a self-hosted Grafana instance. Many teams run Grafana inside a VPC with no public endpoint to visualize metrics, logs, and traces. With the built-in Grafana integration and a private connection, you can give the agent read-only access to your dashboards, alerts, and data sources, the same observability data your on-call engineers rely on during incidents.

AWS DevOps Agent provides a dedicated Grafana integration that hosts the official open-source Grafana MCP server on your behalf, so you don’t need to deploy or manage any MCP server infrastructure yourself. The integration supports Grafana Cloud, Grafana Enterprise, and self-hosted Grafana OSS (v9.1 and later).

For self-hosted instances that aren’t publicly accessible, pair the Grafana integration with a private connection so the agent can reach your instance over the private network.

Step 1: Create a Grafana service account

In your Grafana instance, create a service account with Viewer role permissions. This gives AWS DevOps Agent read-only access to your dashboards, alert rules, and data sources. Generate an access token for the service account and save it for the next step.

Step 2: Create a private connection to your Grafana instance

Because your Grafana instance runs in a private subnet, create a private connection first so that AWS DevOps Agent can reach it.

Using the AWS Console:

Follow the steps in the Create a private connection section of this blog, using your Grafana instance’s address as the host address.

Using the AWS CLI:

aws devops-agent create-private-connection \
    --name grafana-connection \
    --mode '{
        "serviceManaged": {
            "hostAddress": "grafana.internal.example.com",
            "resourceGatewayConfig": {
                "create": {
                    "vpcId": "vpc-0123456789abcdef0",
                    "subnetIds": [
                        "subnet-0123456789abcdef0",
                        "subnet-0123456789abcdef1"
                    ],
                    "portRanges": ["443"]
                }
            }
        }
    }'

Wait for the connection status to reach Completed state before proceeding.

Step 3: Add the Grafana integration to your Agent Space

Register and associate the Grafana service so the agent knows how to connect.

Using the AWS Console:

  1. Open the AWS DevOps Agent console and navigate to your Agent Space.
  2. In the Integrations section, choose Add integration.
  3. Select the Grafana tile.
  4. Enter your Grafana instance URL (for example, https://grafana.internal.example.com).
  5. Paste the service account access token you generated in Step 1.
  6. Choose Save.

Using the AWS CLI:

First, register the Grafana service with your instance URL and service account token:

aws devops-agent register-service \
    --service mcpservergrafana \
    --private-connection-name grafana-connection \
    --service-details '{
        "mcpservergrafana": {
            "name": "grafana",
            "endpoint": "https://grafana.internal.example.com",
            "authorizationConfig": {
                "bearerToken": {
                    "tokenName": "grafana-sa-token",
                    "tokenValue": "<SERVICE_ACCOUNT_TOKEN>"
                }
            }
        }
    }' \
    --region us-east-1

Save the returned serviceId, then associate the Grafana service with your Agent Space:

aws devops-agent associate-service \
    --agent-space-id <AGENT_SPACE_ID> \
    --service-id <SERVICE_ID> \
    --configuration '{
        "mcpservergrafana": {
            "endpoint": "https://grafana.internal.example.com"
        }
    }' \
    --region us-east-1

The response includes webhook information that you can use in Step 4 to configure alert notifications. The agent routes traffic to your Grafana instance through the private connection you created in Step 2, matching on the host address.

To verify the integration, start a chat in your Agent Space and ask the agent something like “Summarize my recent Grafana alerts.” If the agent returns alert data from your dashboards, the private connection and Grafana integration are both working.

Step 4: Enable Grafana webhook notifications (optional)

To let AWS DevOps Agent automatically start investigations when Grafana fires an alert, configure a webhook contact point in your Grafana instance. Point the webhook to your Agent Space’s webhook URL and provide the authentication secret from the integration settings. When an alert triggers, Grafana sends a notification to the agent, which begins an investigation using your Grafana data alongside the other resources in your Agent Space.

For details on the AWS DevOps Agent Grafana integration, and details on how you can also use this integration with AWS Managed Grafana (AMG), check the AWS DevOps Agent Grafana integration documentation.

How DNS resolution and host routing work

When you create a private connection, the host address you provide (for example, my-alb.us-east-1.elb.amazonaws.com) is the DNS name that VPC Lattice resolves to route traffic to your target. This DNS name must be publicly resolvable, even if it resolves to private IP addresses. When you register a service integration and specify an endpoint URL (for example, https://my-grafana.internal.corp), that URL is used for the Host header and Service Name Indicator (SNI) on the TLS connection, it is not used for DNS resolution. This separation means you can point multiple service integrations at the same private connection (and therefore the same target, such as an Application Load Balancer), each with a different endpoint hostname. The target can then use the Host header to route requests to different backends.

For example, you could create a single private connection with your ALB’s DNS name as the host address, and register separate Grafana and MCP server integrations, each with their own endpoint URL. The ALB receives both through the same private connection and uses host-based routing rules to direct each request to the correct target group.

Advanced setup using existing VPC Lattice resources

If your organization already uses Amazon VPC Lattice and manages your own resource configurations, you can create a private connection in self-managed mode. Instead of having AWS DevOps Agent create a resource gateway for you, you provide the Amazon Resource Name (ARN) of an existing resource configuration that points to your target service.

This approach is useful when you:

  • Want full control over the resource gateway and resource configuration lifecycle.
  • Need to share resource configurations across multiple AWS accounts or services.
  • Require VPC Lattice access logs for detailed traffic monitoring.
  • Run a hub-and-spoke network architecture.
  • Require zero trust fine grained access controls for access to your resources or services.

To create a self-managed private connection using the AWS Console, click on Create a new connection, and select Use existing resource configuration, as shown in the following screen capture (figure-8). Then select your existing resource configuration for the drop-down list.

Figure-8: Advanced setting using existing VPC Lattice resources

To create a self-managed private connection with the AWS CLI:

aws devops-agent create-private-connection \
    --name my-advanced-connection \
    --mode '{
        "selfManaged": {
            "resourceConfigurationId": "rcfg-0123456789abcdef0"
        }
    }'

For more details on setting up VPC Lattice resource gateways and resource configurations, see the Amazon VPC Lattice User Guide.

Clean up

To avoid ongoing charges, delete private connections that you no longer need.

Using the AWS Console

  1. Open the AWS DevOps Agent console.
  2. In the navigation pane, choose Capability providers, then choose Private connections.
  3. Select the private connection you want to delete.
  4. Choose Delete.
  5. Confirm the deletion.

Using the AWS CLI

aws devops-agent delete-private-connection \
    --name my-test-private-connection

The response returns a status of DELETE_IN_PROGRESS. AWS DevOps Agent removes the managed resource gateway and ENIs from your VPC. After deletion completes, the connection no longer appears in your list of private connections.

Conclusion

Private connections for AWS DevOps Agent provide a secure, managed way to connect your Agent Space to services running inside your VPC. By using Amazon VPC Lattice, private connections keep all traffic off the public internet while your existing security control remain in place, giving you full control over network access through your own security groups.

To get started, open the AWS DevOps Agent console and create your first private connection. For more details, see Private connections in the AWS DevOps Agent User Guide.

About the authors

Alex Huides

Alexandra Huides

Alexandra Huides is a Principal Networking Specialist Solutions Architect in the AWS Networking Services product team at Amazon Web Services. She focuses on helping customers build and develop networking architectures for highly scalable and resilient AWS environments. Alex is also a public speaker for AWS, and is helping customers adopt IPv6. Outside work, she loves sailing, especially catamarans, traveling, discovering new cultures, running and reading.

Tipu Qureshi

Tipu Qureshi

Tipu Qureshi is a Senior Principal Technologist in AWS Agentic AI, focusing on operational excellence and incident response automation. He works with AWS customers to design resilient, observable cloud applications and autonomous operational systems.

Jordan Merrick

Jordan Merrick

Jordan Merrick is a Senior Software Engineer in AWS Agentic AI, where he works on observability and identity. He builds secure, scalable agentic systems for AWS customers.

Mohak Kohli

Mohak Kohli

Mohak Kohli is a Software Development Engineer at Amazon Web Services. He works across various domains with focus on VPC Lattice and PrivateLink.

A Taxonomy of Cognitive Security

Post Syndicated from Bruce Schneier original https://www.schneier.com/blog/archives/2026/04/a-taxonomy-of-cognitive-security.html

Last week, I listened to a fascinating talk by K. Melton on cognitive security, cognitive hacking, and reality pentesting. The slides from the talk are here, but—even better—Menton has a long essay laying out the basic concepts and ideas.

The whole thing is important and well worth reading, and I hesitate to excerpt. Here’s a taste:

The NeuroCompiler is where raw sensory data gets interpreted before you’re consciously aware of it. It decides what things mean, and it does this fast, automatic, and mostly invisible. It’s also where the majority of cognitive exploits actually land, right in this sweet spot between perception and conscious thought.

This is my term for what Daniel Kahneman called System 1 thinking. If the Sensory Interface is the intake port, the NeuroCompiler is what turns that input into “filtered meaning” before the Mind Kernel ever sees it. It takes raw signal (e.g., photons, sound waves, chemical gradients, pressure) and translates it into something actionable based on binary categories like threat or safe, familiar or novel, trustworthy or suspicious.

The speed is both an evolutionary feature and a modern bug. Processing here is fast enough to get you out of the way of a thrown object before you’ve consciously registered it. But “good enough most of the time” means “predictably wrong some of the time….

A critical architectural feature: the NeuroCompiler can route its output directly back to the Sensory Interface and out as behavior, skipping the conscious awareness of the Mind Kernel entirely. Reflex and startle responses use this mechanism, making this bypass pathway enormously useful for survival. Yet it leaves a wide-open backdoor. If the layer that holds access to skepticism and deliberate evaluation can be bypassed completely, a host of exploits become possible that would otherwise fail.

That’s just one of the five levels Melton talks about: sensory interface, neurocompiler, mind kernel, the mesh, and cultural substrate.

Melton’s taxonomy is compelling, and her parallels to IT systems are fascinating. I have long said that a genius idea is one that’s incredibly obvious once you hear it, but one that no one has said before. This is the first time I’ve heard cognition described in this way.