The people most likely to be hurt by the gap between perceived and actual privacy are the ones who can least afford to find out the difference too late.
This article explains in simple terms how the technology stack for every mobile user works, what each layer exposes, and what people are actually trusting when they hand God Mode to a new MVNO.
Your phone is broadcasting your identity, location and behaviour to multiple companies right now, regardless of which privacy app you’re using.
Encrypted apps like Signal encrypt the content of your messages but your mobile network still sees who you’re communicating with, when, and from where.
Cape, a new mobile virtual network operator in the US, promises to protect the privacy of high risk individuals, but was founded by former Palantir executives who built their careers serving the CIA, NSA and FBI, won their first government contracts before they had a single consumer on their network. They’re operating a system that combines the surveillance capabilities of a carrier like AT&T with the interception capabilities of companies like
SS8 Networks, built specifically to tap into those networks on behalf of intelligence agencies.
The 4 layers almost every privacy app operates inside
To understand what any privacy tool can and can’t protect you from, you need to understand the 4 layers in the technology stack:
- Layer 1: The app. Signal, WhatsApp, Telegram. This is where encryption lives. It protects the content of messages between devices.
- Layer 2: Your phone. A physical device with a permanent hardware identity that broadcasts to every cell tower it passes, regardless of which app you’re using.
- Layer 3: The mobile network. Records your location, device identity and communication patterns continuously and automatically.
- Layer 4: The lawful interception infrastructure. Companies like SS8 Networks that physically tap into mobile networks on behalf of intelligence agencies and law enforcement.
Signal secures Layer 1. Layers 2, 3 and 4 remain fully visible to anyone with the right access and the right authority.
What Signal protects and what others still see
Signal is a well built product and the Signal Protocol is strong. When you send a message on Signal, the content is encrypted between your device and the recipient’s. Nobody in the middle can read it, unless the CIA has a backdoor.
But think of your phone number as a PO box address. Every message you send or receive is logged against that address, along with the address of every person you communicate with, the time of every exchange and how often. Your phone connects to cell towers continuously, recording your location every time it does. Your carrier logs all of this regardless of which app you’re using, and retains it for years.
So even though nobody can read your messages, your carrier knows your PO box address, every address you’ve ever exchanged messages with, exactly when those exchanges happened, how often, and where you were each time. A regular carrier collects more than 100 data points on each customer. When that dataset is combined with your communication patterns, relationships and location history, it can identify who you are, map your social network, track your movements and predict who you’ll meet before you meet them, without a single word of your messages ever being read. With modern AI that analysis takes seconds and alerts can be automated to notify whoever has access the moment something triggers their interest. Signal locks the letter inside the envelope but it has no control over what’s written on the outside, and the postal system records every delivery.
The infrastructure built to tap into what the network sees
In the late 1990s I was seconded from ADC Metrica, a specialist telecoms testing and measurement consultancy, to audit the technical operations of NewNet in Connecticut on behalf of ADC Telecommunications after its acquisition. My role was to evaluate the performance of the people responsible for all of their technology development and testing, spanning 12 development managers and 2 test managers along with their internal teams and outsourced partners, and to make recommendations on what to do with them.
NewNet was a powerhouse in SS7 and SMS technology that provided Lawful Interception services, the infrastructure that makes it possible for law enforcement in the US to intercept and monitor communications at scale. They offered me a full time role to manage the combined operation but Connecticut wasn’t appealing to me during my 20s.
What I didn’t know at the time was that the technology I was auditing would become the foundation for something much larger. When ADC later divested that division, SS8 Networks acquired it and built it into one of the world’s leading lawful intelligence platforms, trusted by 6 of the world’s largest intelligence agencies. SS8’s systems physically tap into mobile and IP networks to enable real time interception of voice, data and metadata.
The companies that analyse what interception captures
Palantir is arguably the most powerful company in the world when it comes to what it knows about people and what it can do with that knowledge. Its clients include the CIA, NSA, FBI and the UK Ministry of Defence. Its Gotham and Foundry platforms process and visualise the kind of communications data that systems like SS8’s are built to capture. SS8 captures. Palantir analyses. Together they serve the same intelligence and defence ecosystems, performing complementary functions inside the same surveillance infrastructure, and the capabilities that combination unlocks are not fully visible to the public.
The company that combines all of these layers under one roof
Cape is a 4 year old Mobile Virtual Network Operator (MVNO) that has raised $191 million and is valued at $800 million, backed by Andreessen Horowitz and Bain Capital Ventures, two of the most prominent investors in US defence and intelligence technology, including Palantir and Anduril.
SS8 was built to tap into networks owned by carriers like AT&T and extract visibility over who is communicating with whom, when and from where. Cape isn’t just building its own core network with unknown partners, it’s building tech that has significantly more invasive capabilities. Based on what Cape says it does and how, it has the potential to make AT&T and SS8 redundant for anyone seeking access to a subscriber’s data.
Cape doesn’t just collect the same kind of data as AT&T, it collects more. Its privacy features, IMSI rotation, Network Lock, encrypted voicemail and SIM swap protection, require Cape to see, process and retain far more data about each subscriber than any standard carrier ever collects. Network Lock requires continuous GPS access. IMSI rotation requires a real time mapping table linking every rotating identity back to the real subscriber.
And buried in Cape’s privacy policy, not on their website, is a capability that allows Cape or their network partners to scan the content of SMS messages if they suspect spam or a scam campaign. There’s no opt in or opt out, which for a privacy network built specifically for the highest risk individuals is a significant omission.
Cape also requires subscribers to use their mandatory app, for reasons that don’t live up to my scrutiny, and offers a customised phone. Every phone contains a baseband processor, a separate chip that runs its own closed-source firmware and manages all communication with the mobile network. It operates independently of the main operating system, sitting beneath any privacy controls the customer can see or configure. A manufacturer with full control over both the hardware and that firmware could embed capabilities that would be extraordinarily difficult for even expert security researchers to detect. The
BADBOX operation, in which tens of thousands of consumer Android devices were shipped with firmware backdoors that went undetected at scale, demonstrated that this isn’t theoretical. No standard carrier has access to the hardware layer of a subscriber’s device. Only the device manufacturer does. Cape is both.
All of this means an intelligence agency with an existing Palantir relationship wouldn’t need to go beyond it if Cape operates as an undisclosed partner or functions within Palantir’s ecosystem. The data collection, routing, monitoring, analysis and delivery to authorised parties can now flow through a single private tech company.
If you were designing a system that gave a single organisation complete visibility over the communications of the highest risk individuals in the United States, it would look a great deal like what Cape has built.
Who built Cape and why their background matters
If you’re going to trust a company with God Mode to everything covered in this article, you need to know who you’re trusting.
All 4 founders come from what’s known as the “Palantir Mafia”.
- John Doyle, Cape’s CEO, was Head of National Security at Palantir, directing the division responsible for delivering intelligence and data analytics to the CIA, NSA, FBI and DoD.
- Nicholas Espinoza, Head of R&D, his role at Palantir involved identifying cellular network vulnerabilities and building analysis platforms for the intelligence community.
- Stephen Dowhy, Head of Engineering, comes from Anduril Industries
- David Dunn, Chief Architect, also from Palantir.
Other members of staff and advisors come from agencies including the CIA, FBI and Homeland Security.
Cape’s own press materials describe the company as founded by experts in national security, and Doyle himself describes their technology as originally designed alongside national security professionals. Their mission, to ensure communication is private, secure and resilient, mirrors Palantir’s own national security mandate almost word for word. One of Cape’s investors has confirmed that the most recent National Defense Authorization Act includes explicit language requiring federal agencies to adopt the precise capability Cape is built around: rotating and obfuscating persistent device identifiers. Cape lists its customers as government, enterprises and consumers, in that order.
In the context of a company like Cape, the ability to rotate these identifiers is controlled entirely by its technology. The high-risk subscriber is trading limited tracking by regulated carriers for the most invasive tracking I’ve seen, by a tech company that holds the master key to those rotating identities.
Cape’s relationship with the US government and what it reveals
Within a year of founding, with no consumer product, no subscribers and no track record as a carrier, Cape secured government contracts including a pilot with the US Navy in Guam to secure communications for military personnel, contractors and their families. Building and deploying a carrier grade mobile network to a standard that satisfies US military requirements doesn’t happen in 12 months from a standing start. No serious infrastructure company achieves product market fit that quickly, let alone one with a significant technology stack that wins sensitive government contracts on the strength of a promise.
The most credible explanation is that the groundwork was laid before Cape existed as a company, while they were still working at Palantir, where they already had the relationships, the knowledge and the access that only comes from years inside the national security apparatus.
Cape has raised its funding from the same investor ecosystem as Palantir and Anduril. And yet there’s no public comment from Alex Karp or Peter Thiel about Cape. Cape’s CEO has been defending claims on public forums like
Hacker News that Cape is a “
honeypot” by acknowledging they have to earn trust.
What makes the Guam deployment particularly striking is its sequence. Typically a consumer technology company proves its product in the public market before government agencies trust it with sensitive operations. Cape did it the other way around. The US military was their first customer. The consumer offering came after. Was Guam a military pilot later adapted for consumers, or was it always the proving ground for a product designed to be sold to the most privacy conscious, highest risk individuals in the United States?
Whether that’s the intent is not something anyone outside Cape can know. But it’s what the architecture makes possible, and for high risk individuals, what an architecture makes possible matters more than what a company promises.
The government trusted Cape before the public had heard of them, and that trust came from who the founders were and what they had already built, not from anything Cape had publicly demonstrated. Before you hand over God Mode to your communications, it’s worth asking whether your reasons for trusting them are the same as the government’s.
Some of their core staff and advisors include former senior officials from the FBI and CIA, including a former COO of the CIA. These are architects of the same God Mode surveillance systems that high risk individuals are trying to hide from. The question of why they’re now building the network those individuals are supposed to hide in is one Cape has never been asked.
30 years of watching the same pattern
Across more than 30 years working inside telecoms infrastructure, consumer technology and cybersecurity, I’ve watched the same pattern repeat. The tools get more sophisticated, the marketing gets more reassuring, and the underlying exposure stays largely the same because the architecture beneath it all hasn’t fundamentally changed. Before you sign up for the next big thing in privacy, look at who’s holding the keys and where they learned to use them.
If I’ve got anything wrong, please tell me. I’d rather be corrected publicly than leave an inaccuracy standing. Everything in this article reflects my own analysis and opinions based on publicly available information. It is not a formal review, assessment or position from MetaCert.