Volt Typhoon botnet rebounds with 1,500 routers while OpenAI deletes China-linked influence accounts

A China-nexus botnet that the FBI disrupted in January 2024 is back in a new form. Lumen’s Black Lotus Labs reports that the Volt Typhoon KV cluster became largely defunct after the takedown, but the JDY cluster remains active and has surged to more than 1,500 compromised routers and IoT devices.

That detail matters because it shows the attack work did not fully stop. The earlier KV-botnet had four clusters, with KV used as a covert data transfer network and JDY used for scanning and reconnaissance. After law enforcement killed KV, JDY kept going, and Lumen says activity since the takedown shows a focus on finding vulnerable infrastructure shortly after public vulnerability disclosures, with reconnaissance rapidly operationalized by China-nexus APT actors.

So what exactly is the threat here for executives and operators? It is not just “we got hacked.” It is the mechanics of how long-running intrusion campaigns use patience, automation, and network sprawl. Routers and IoT devices are often overlooked in security programs, partly because they sit at the edges of enterprise networks and partly because owners do not always treat them like true production systems. If you have any presence in critical infrastructure adjacent roles, government supply chains, telecom, or organizations connected to high-value networks, JDY’s continued scanning and reconnaissance means your perimeter is still being mapped.

Lumen’s report also flags where the reconnaissance effort has landed. It says the targeted focus has been observed across a range of sectors, with the US military and associated entities as the most prominent. For companies serving those ecosystems, this is second-order risk: even if your organization is not the target you think you are, your vendors, access paths, and shared connectivity can become the path of least resistance for the next stage.

Now layer in the other half of the headline: the influence operation playbook using AI. OpenAI said it banned ChatGPT accounts likely originating from China after they used OpenAI’s models to generate content for covert operations. Ben Nimmo, principal investigator on OpenAI’s Intelligence and Investigations team, told reporters that “neither campaign appears to have gained much authentic engagement,” but he called the attempt important for what it reveals about the intentions of influence operators from China and the narratives they were testing.

The first set of banned accounts used ChatGPT to produce social media content and images for an operation claiming datacenters and AI applications increase electricity demand and cause higher costs for ordinary Americans. The report says the prompts included examples like comic strips about a power grid operator’s capacity auction prices based on reporting from a legitimate regional paper. The goal was framing: peak electricity demand is driven by data centers and AI, and the resulting capacity prices are passed to households. OpenAI suspects these operators are part of a social-media team at a private Chinese tech company that provides services for Chinese provincial-level government clients.

The second cluster also likely originated in China. It used OpenAI’s models to write comments and draw political cartoons criticizing US tech policies and tariffs. Nimmo said the operators specified in prompts that outputs should not include cartoons of Xi Jinping, and should only include President Trump. According to OpenAI, these accounts used simplified Chinese and VPNs to access AI systems, and they went beyond posting by using ChatGPT to edit work reports and design social media monitoring systems. That is a key operational signal: even when public-facing output does not take off, the back-end capability building continues.

OpenAI notes it is not the first time actors in China tried to generate ideas for social media monitoring. In February, OpenAI said it banned ChatGPT accounts believed to be linked to Chinese government entities attempting to use AI models to surveil individuals and social media accounts. If AI tools are constrained, the broader pattern is that influence and intelligence operators shift tactics rather than stop. That brings us to the parallel US Justice Department case described in the source: fake websites and job offers promising cash remain on the table.

On Wednesday, the feds said they obtained a warrant for and seized 13 fake consulting company websites used to target US persons, including current and former security clearance holders. DOJ lists domains including centrikglobalconsulting.com, rightinfoconsult.com, finnaclevesperconsulting.com, cydfconsulting.com, pulsewaveglobal.com, catalystglobalsolutions.com, thehorizzen.com, geoindopacific.com, gpf-ina.org, safesec-group.com, thetruthinfo.com, Vandercons.com, and gulfpeace.org. Since November 2023, these sites and associated job postings on social media, LinkedIn, and other hiring platforms advertised “consulting” jobs such as “Senior Analyst” and “International Affairs Consultant” positions.

DOJ alleges PRC operatives used the sites and listings to recruit applicants and bribe them for sensitive information. Court documents quoted in the source say recruiters pressured candidates to share confidential information and reports from “insider sources” in violation of official duties, with payments allegedly made using online accounts under fictitious individuals and cryptocurrency to conceal identities and payment sources. The message for leaders is blunt: cyber intrusion, narrative influence, and human recruitment scams can run side by side, each reinforcing the other by widening access, increasing uncertainty, and probing for weaknesses in systems and people.

For boards and executives, this combination is a governance wake-up call. Lumen recommends enterprises implement CISA and NCSC guidance for mitigating Volt Typhoon activity. OpenAI warns that influence attempts may not have succeeded in engagement, but they still test narratives and build capabilities. DOJ shows the recruitment layer can go after the human decision-makers and clearance holders that most companies underestimate. In other words, this is not a single incident. It is a pressure campaign across networks, platforms, and identities, and JDY’s rebound means “disrupted” did not mean “resolved.”