CAN YOU HEAR ME NOW, BEIJING? We were interested to learn that the Pentagon uses more than 568,000 wireless phone lines that are covered by a $2.67 billion contract. What it apparently didn't have though, was a reliable way to make sure the phone companies that were operating them, were issuing safeguards to protect them. The Daily Beast reports that a new DOW Inspector General's report found that officials failed to adequately verify carriers’ security protections before renewing their contracts. (Here’s why this matters: the vulnerabilities could allow foreign spies to track troops, identify their contacts and potentially map sensitive locations, including command posts and safe houses.) This makes us nervous because we already know that Chinese hackers have already penetrated major U.S. telecommunications networks (Salt Typhoon). In 2024, a bipartisan pair of senators Eric Schmitt (R-Mo.) and Ron Wyden (D-Ore.) urged an investigation and the Pentagon subsequently added requirements for independent security assessments. But the inspector general found that officials still lacked a proper process for reviewing the results. It’s important to note that the audit didn’t establish that foreign intelligence services had actually compromised military accounts that were covered by the contract. The Pentagon says it is strengthening protections.
TOILET PAPER TRAIL: The FBI calls it a counterintelligence priority. The charge sheet calls it eavesdropping. DOJ says OpenAI engineer Abdulrahman Alfozan has been charged with hiding a listening device inside a roll of toilet paper in the company’s executive conference room in San Francisco. According to the complaint, as reported by Bleeping Computer, that room hosted meetings on research, product plans and government engagements. The device allegedly went in around September 28 and was discovered two days later. Security staff reportedly found a second one in a bag believed to belong to Alfozan. FBI Director Kash Patel and Attorney General Todd Blanche cast the case as a national security matter, citing the protection of American “superintelligence” technology. The actual charge, intercepting oral communications, carries a maximum of five years in prison. Alfozan is a Saudi national, but prosecutors have not alleged that he was working for a government. OpenAI says it has seen no sign he took model weights, source code or customer data. The AI angle cuts both ways. OpenAI says its own models analyzed the device and identified who planted it. Prosecutors say a chatbot account linked to Alfozan’s spouse had asked how to build and conceal recording devices. Seems to us, creating one and wrapping it in toilet paper is not exactly “super intelligence.” Alfozan didn’t enter a plea during his first court appearance, so we may have to wait to hear what he says about all of this.
HONEY, DON’T: OpenAI’s new security curriculum opens with the oldest chapter in the book. Shortly after the FBI arrested OpenAI engineer Abdulrahman Alfozan for allegedly hiding a recording device in an executive conference room, there are reports that the company has moved on to the next unit of the counterintelligence syllabus. That chapter is on seduction. According to The Wall Street Journal, OpenAI’s internal training now warns staff about “honeypots,” the trade term used for agents who use romance or sex to get close to a target. The case study apparently comes from Moscow. Imagine a blonde woman shows up with champagne at a U.S. diplomat’s hotel room. When he declines, his would-be handlers sent a redhead, then they sent a young man as they try to figure out which bait he likes. Business Insider reports that the tech industry’s response was to start ranking engineers on a 1-to-10 scale to estimate each one’s personal threat level. The New York Post says OpenAI’s message is that if you are getting hit on – it is probably by a spy. Or as one X poster said, ‘Remember, you are not that good looking and not that funny.’
CLICK NOTHING, FEAR EVERYTHING: We’ve all suspected that online ads know a little too much about us. It turns out, they may also be spying on us. Reuters reports that Israeli ad company Discover-Tech turned online advertising into a delivery system for surveillance and spyware, and gave Israeli spyware maker Candiru access to its platforms. (The U.S. Commerce Department blacklisted Candiru in 2021.) According to the University of Toronto’s Citizen Lab, Discover-Tech ads ran on more than 120,000 websites and 75,000 apps, usually hawking ordinary stuff like phones and earbuds. The catch: nobody had to click. Simply loading an ad could help identify and track a user and, in some cases, silently deliver spyware. Citizen Lab said it found at least 45 Windows computers at Al Jazeera that were infected with Candiru spyware in 2024 and 2025. Researchers also say they spotted the ads on networks tied to NATO, the Pentagon, the UN and the International Criminal Court, though seeing an ad didn’t necessarily mean a device was compromised. Ad-security firm Confiant says Discover-Tech served at least 10 million ads a day, and co-founder Jerome Dangu estimates “hundreds of millions of people” were touched by its tracking. For a company in the surveillance business, Discover-Tech was remarkably easy to watch. TechCrunch reports it left source code exposed on its login pages and let anyone download user login records, including some belonging to Candiru employees, without a password. Google has since booted the company from its ad platforms. Reuters reportedly reached out to Discover-Tech for a comment but say the company didn’t respond to messages left via its contact form. A LinkedIn message and emails sent to its chief executive were also reportedly not returned.
HAPPY BIRTHDAY TO SOMEBODY’S NAVY: The U.S. Navy turned 251 on October 13th, which means it’s time for our favorite annual sport: politicians saluting the American fleet with somebody else’s vessels. The Wall Street Journal’s Dave Brown once again rounded up the entries on X. Georgia Governor Brian Kemp led off with a “Happy Birthday US Navy” graphic featuring a column of Indian warships. North Carolina Congressman Tim Moore used what looks like the same Indian formation, stamped “251 Years.” Perhaps most embarrassing, Ohio State Representative Mike Dovilla said he was proud of his service in naval intelligence and posted a photo of Russian warships. Dovilla is a commander in the Navy Reserve and a career intelligence officer, according to his Ohio House biography. Jeffrey Dinise, who runs the Border Patrol’s Miami Sector and is responsible for more than 1,200 miles of coastline, picked a Russian Kirov-class battlecruiser to cosplay a USN ship. Another post used a destroyer that looks official until you zoom in and notice it was apparently made by AI. A year ago, a Navy SEAL museum marked the 250th with a Russian cruiser. Memo to politicians: when it becomes time to celebrate the 252nd birthday – ask the U.S. Navy to give you some real photos of their ships. We are pretty sure they have some.
UPLINK DOWN IN FLAMES: Two Turkish nationals have been detained in Germany over an alleged arson attack on a Starlink ground station in Poland. AP reports that Polish officials think Russian intelligence was behind the September 23 fire at Wola Krobowska, about 25 miles south of Warsaw. The station is run by the Polish state telecom Exatel and helps route satellite internet to Ukraine. An organization called Notes from Poland reports the two were picked up Monday in Nürtingen, near Stuttgart, at Warsaw’s request. “We have no doubt that Russian intelligence services are behind this act of sabotage,” said Tomasz Siemoniak, the minister who oversees Poland’s security services. He added that Moscow is “increasingly using people connected to the criminal world.” Polish police say they have recordings of the suspects, plus their fingerprints and DNA. The likely charge is sabotage on behalf of a foreign intelligence service, which carries a minimum of 10 years. They’re set to go before a judge soon.
ARTIFICIAL INTELLIGENCE, REAL GERMS: MI5 Director General Ken McCallum has a new worry: terrorists using AI to build biological weapons and maybe even using pathogens that don’t even exist yet. In his annual threat update, McCallum said AI could put dangerous knowledge in the hands of people who otherwise wouldn’t have it. MI5 has set up an AI Threat Assessment Centre with the UK’s AI Security Institute to track how hostile actors misuse the technology. “Human history has given us no experience of facing an actor which is smarter than we are,” he warned (presumably referring to the AI and not the terrorists.) The older threats are still growing. Newshub reports that MI5’s state-threat caseload has quadrupled in the past decade. And check this out: in the past year alone, officials say Iran-linked operations led to more than 50 arrests in Britain. McCallum said Tehran’s targets now go beyond dissidents and Jewish communities to include military installations. MI5’s counterintelligence budget now tops its Cold War levels.
SPY HELP WANTED: PIZZA BAG PROVIDED: A Russia-based cyber-extortion gang apparently has ambitions well beyond stealing your bank account password. According to The Record, leaked internal chats from an outfit known as the “Silent Ransom Group” show members plotting to send Russian-speaking recruits into U.S. law firms, kidnap executives and even gather intelligence on America’s submarines. Despite the ambitions, the recruiting was decidedly low-tech. Telegram job ads offered paid gigs as nightclub promoters, couriers and security workers. Successful applicants could end up posing as IT technicians and copying confidential files onto thumb drives, a tactic the FBI has already warned about. This gets even more interesting though, with one alleged scheme buying pizza-delivery uniforms and insulating bags to get past reception. A separate channel discussed recruiting U.S. sailors near naval bases to track the movements of nuclear-armed subs. Nothing in the chats shows that plan went anywhere though. The archive mixes real extortion records with boasting and brainstorming, and a Blockchain analysis firm has corroborated some of the crypto payments, though not the whole trove. The gang’s apparent leader estimated that only one in ten recruits was usable, calling it the operation’s “conversion rate.” Spoken like a true human resources manager.
WHAT’S THE DEAL WITH THE RANSOM NEGOTIATOR? Edward Dubrovsky made a career helping organizations deal with ransomware gangs. Now the Canadian cybersecurity executive is facing U.S. federal extortion charges. According to CyberScoop, Dubrovsky, 54, was arrested in Pennsylvania on October 8. He is the former chief operating officer and co-founder of CYPFER (no relation to us at “Cipher”) a firm specializing in helping victims of cyberattacks, including negotiating with the criminals. Now, Dubrovsky is charged with conspiring to threaten the confidentiality of information to extort money, in violation of federal computer fraud law, and conspiring to commit Hobbs Act extortion, which involves using threats to affect interstate commerce. Other details of the case remain sealed. The New York Times and CyberScoop suggest that the case may be connected to the FBI's investigation of ShinyHunters, the hacking group that claimed responsibility for breaching the bureau's jobs portal and stealing sensitive personal information about employees and their families. The FBI is being a bit cryptic about the matter – judging by Director Kash Patel’s X post. Less than three weeks ago, Dubrovsky published a book called “Cyber Extortion Strategic Response.” So far, we have seen no response, strategic or tactical - from him or his attorney.
FOWL PLAY COOP D’ETAT: An Israeli poultry farmer reportedly went to check on his chickens and discovered that someone had turned his security camera into a spying mechanism. According to The Wall Street Journal, the farmer in northern Israel opened his webcam app during the war with Iran and found that the camera's language had been switched to Arabic. Israel's National Cyber Directorate investigated and discovered that hackers had commandeered the Chinese-made device. Their interest wasn't exactly poultry production. The camera had a view of Israeli security-force movements near the Lebanese border, making the chicken coop an unlikely intelligence collection post. Officials helped the farmer secure the camera and change its viewing angle. The incident is part of a much larger problem: Israeli authorities say Iranian-linked hackers have compromised hundreds of sensitive Israeli cameras, while Russia and Ukraine have also reportedly been hijacking civilian webcams to track military movements and help identify targets in their own war. It seems like cheap (or maybe cheep) intelligence has come home to roost. (We couldn’t resist.)
A MOLE IN MOLESWORTH? Two Latvians walk into a NATO intelligence hub. Police are still working on the punchline. The Associated Press reports that British counterterrorism police are trying to figure out why two Latvian citizens living in the U.K. entered RAF Molesworth, an installation in Cambridgeshire that hosts American and NATO intelligence operations covering Europe and Africa. On the night of Oct. 7, a security patrol reportedly spotted an abandoned car outside the base and then discovered a small hole in the perimeter fence. The men, ages 32 and 36, were arrested early on Oct. 8, according to press accounts. They’re being held on suspicion of trespass and entering a prohibited place “for a purpose prejudicial” to U.K. security, an offense under Britain’s 2023 National Security Act. Commander Helen Flanagan, head of Counter Terrorism Policing London, said police are “very aware of the recent heightened interest and concern around military sites.” She added there is “nothing to suggest” a link to RAF Fairford, where seven men were arrested after a security scare on Sept. 27.
Got news to share? Drop us a note: Editor@thecipherbrief.com
Read more expert-driven national security insights exclusively in www.thecipherbrief.com.
