Sample report written from an AI-Dossier by a fictional client
Seven pages from an invented case, marked Sample, fictional data. Read it here or download the PDF. The full sample dossier needs a short form.
A short, curated set of threat briefs, release notes and fact sheets. Every item carries its date.
Editions, evidence types, the detection engines, dated SectraTEK Intelligence totals, support and security at a glance, on one page you can print.
Seven pages from an invented case, marked Sample, fictional data. Read it here or download the PDF. The full sample dossier needs a short form.
Apple has warned users in more than 150 countries since 2021, most recently in 110 on 13 August 2026. A notification is a high-confidence warning, not proof; Lockdown Mode is strong, but Apple says only that it knows of no successful attack.
According to Citizen Lab, an Israel-based ad-tech firm let customers aim an advertisement at one device, and servers that Citizen Lab links to a spyware vendor then profiled that device. An ad impression alone, the lab stresses, establishes neither targeting nor infection.
An advanced Android acquisition goes beyond a bug report with a fuller app inventory, a snapshot of running processes and, when included, a record of network use. SectraTEK Research Lab explains what SectraTEK reads in it and how findings reach the client.
In 2026 Calif says AI models helped it build a worm demonstration, test a Mac memory defense and stage a jailbreak in Paris. The details are withheld; what can be checked is the gap between a fix and a public crash file.
The most inventive Android threats of 2026 do not break the phone; they persuade its owner. A staged call, a sideloaded remote-control app and a card relay over NFC show how far a conversation can go.
An Android bug report is the diagnostic bundle a phone produces about its own state, logs and settings. SectraTEK Research Lab explains what SectraTEK examines in it, where it falls short of a full extraction, and how the findings are packaged as a prioritized, evidence-referenced AI-Dossier.
An Android full-filesystem extraction holds the apps, permissions and records a handset kept about its own life. SectraTEK turns that record into rated, evidence-referenced findings a client can act on.
Google marks only a few Android bulletin entries each year as possibly exploited, and it chooses its words with care. Reading that wording, and the patch level on a phone, tells you more than any zero-day headline.
Google described a kit of 23 exploits in five chains for iPhones on iOS 13 to 17.2.1, hidden behind web pages. Within a year, according to Google, it passed from a surveillance vendor's customer to a suspected espionage group to criminals stealing cryptocurrency.
Google, Lookout and iVerify each documented a chain of six iPhone flaws, written entirely in JavaScript, that started when a phone visited a page. Several operators used it on iOS 18, it deleted crash reports behind it, and Google and Lookout agree iOS 26.3 and 18.7.3 fall outside it.
For years one small system file preserved traces of older spyware, until early iOS 26 releases began overwriting it at every restart. Newer implants go further: they delete crash reports, vanish on reboot or leave when analysts appear.
Two weeks after Google described DarkSword, a newer build was reported on a public code site, and iVerify says variants followed. Our assessment separates what is confirmed from one vendor's claims, and says who stays at risk.
At a two-day offensive-security conference in Paris, researchers showed an iPhone restarting into a root shell and a run of Android and boot-level research. Almost none of it has been released, and that gap is the story.
A study of 82 Android stalkerware builds and an independent test of 13 security products explain why these apps persist: they ask for accessibility rights, restart themselves, hide behind generic labels and split into a compliant app and a hidden one. The same test shows why a warning on the victim's screen can itself be dangerous.
An iOS backup is a copy of a phone's apps, data and settings made through a computer. SectraTEK Research Lab explains how SectraTEK examines it for known-bad indicators, risky apps and exposed configuration, and where a backup falls short.
A full-filesystem extraction is the most complete copy of an iPhone that an examination can start from. SectraTEK Research Lab explains what SectraTEK reads in it, what even this evidence cannot show, and how a long examination is packaged for a short, defensible report.
A sysdiagnose is the diagnostic archive an iPhone writes about itself: logs, crash reports and system state. SectraTEK Research Lab explains how SectraTEK examines it for permissions, known-bad indicators and exposure, and packages what it finds as a graded, evidence-referenced AI-Dossier.
Apple's Memory Integrity Enforcement, built into the newest iPhone chips, aims to make the class of bug behind many exploit chains harder to use. Apple says it cut off six real chains it rebuilt; outside testers have claimed progress on a Mac, and no published evidence confirms a bypass on an iPhone.
Messaging apps hold more of an owner's life than any other app, and attackers have noticed: group invitations, linked-device QR codes, link previews and one demonstration that needs only an incoming call. The evidence behind each route is uneven, and this article grades it route by route.
Ninety zero-days were exploited in the wild in 2025 and 15 of them hit mobile operating systems, according to Google. The headline numbers are all floors drawn from someone's telemetry: here is what each measures and what no one has measured.
A network capture shows who a phone talked to, when and how often, even when every byte is encrypted. SectraTEK reads that metadata, ranks what stands out and carries it, in an AI-Dossier, to the report a client can act on.
A zero-click attack needs nothing from the phone's owner; a one-click attack needs a single tap, open or scan. Cases documented from 2021 to 2026 show how each works, why so little remains afterward, and how rare these attacks are.
Some Android phones, tablets and streaming boxes carry malware before the first boot. Researchers describe system-level apps that cannot be removed and firmware backdoors used for ad fraud, with the caution that a supply-chain link is not proof of intent.
Confirming that an iPhone ran mercenary spyware takes more than a warning: published cases rest on device logs, server fingerprints and message accounts. That standard explains why so many phones stay 'inconclusive', and why a negative is not an all-clear.
Courts, trade rules and sanctions are now part of the story of commercial spyware. A US court enjoined one vendor from attacking WhatsApp, the Commerce Department's entity list and Treasury sanctions have shifted, and a voluntary international code is being drafted. Each item here carries its date and its status.
Stalkerware is bought by people who know the victim, not built by governments. The counts are floors from one vendor's users, and the first rule for anyone who finds it is safety, not removal.
The first code an iPhone runs is etched into silicon, so a flaw there cannot be patched. One was published for A12 and A13 chips in June, and a Hexacon talk described a broader approach; both need the phone in hand.
On 16 October the research firm Calif showed an iPhone 18 Pro on iOS 27.0.1 restarting, unlocking with its normal passcode and offering a root shell again, according to press reports. Calif has released no tool, bug list or date, and Apple has not commented.
Phones decode photos, audio and fonts constantly, often before the owner looks. A run of 2025 and 2026 cases shows why those decoders are targets, how some chains used a message with no tap, and how researchers keep finding the flaws first.
Some of the most effective PDF attacks contain no exploit at all. A picture of a QR code sends the reader from a monitored computer to an unmonitored phone, and a PDF reader in an app store can turn out to be an installer for something else.
A phone is not only attacked over the air. A cable, a lookalike charger or an extraction device can talk to the kernel, and the tools examiners connect to hostile phones parse hostile data. Both facts argue for isolation, updates and a second opinion.
In September Apple fixed a flaw in the graphics engine that renders PDFs and said it may have been used in an extremely sophisticated attack on specific people. Here is what is known about that flaw, what researchers published afterward, and what nobody knows about how it was delivered.
Mobile compromise rarely announces itself, yet the documented attack routes start in the same four places: the browser, the sensors, the documents that arrive and the traffic nobody watches. Independent research shows why those places hold the evidence, and what they cannot prove.
Public jailbreaks still exist, but only for older chips and older software; for current iPhones none has been released. A jailbroken phone is a different kind of evidence source, and so is a phone whose owner was told that one is on offer.
Editions, evidence types, the detection engines, dated SectraTEK Intelligence totals, support and security at a glance, on one page you can print.
The SectraTEK film, with chapters that take you straight to any part. Two short clips are on the Videos page.
Plain-language definitions of the terms you meet in SectraTEK and in mobile threat examinations.
The documents and answers a procurement team asks for, in one place: security overview, accessibility, license terms, pricing and how to request a quote.
The complete AI dossier of the same invented case: the examination data behind every finding, with the evidence integrity record. Marked Sample, fictional data.
A match is a lead with context, not a verdict. How to read the family, the campaign and the evidence around a match before you decide what it means.
Guided collection, a clear result from every engine, a dossier that waits for your approval and evidence that stays where it is. What is new, in plain language.
Profiles are how organizations manage iPhones, and a place an unwanted change can hide. How to tell an expected profile from one that deserves a second look.
A few permissions give an app a great deal of reach. Which to check first on an Android phone, and how to read them against what the app is for.
Try a shorter phrase, or browse by evidence type.
It plays only when you press play.
