An increasing number of persons are utilizing AI like a search engine – 42 p.c of U.S. adults now use AI chatbots to seek for data – and that shift is exposing a structural vulnerability that adversaries are exploiting to seed propaganda. In sensible phrases, this can be a drawback of Generative Engine Optimization (GEO) – the deliberate effort to form digital content material in order that AI chatbots take up and repeat it.
The analysis to this point factors to knowledge voids — the thinly lined matters the place there is not a lot credible data to start with — because the weak spot. This contains breaking information or new materials that has not but had time to build up the alerts that will flag it as low authority.
AI can course of much more data than any human can, however there’s an inherent trade-off in outsourcing the curation of knowledge to an AI abstract. Within the search period, customers had been uncovered to supply materials and evaluated credibility for themselves. Now AI does that work, and analysis exhibits the massive majority of AI queries finish with out a click on.
That is each a enterprise concern and a nationwide safety concern. The clearest instance on the buyer aspect is Apollo-9. In a Chinese language state TV investigation, researchers used a software referred to as Liqing to flood the online with pretend evaluations and rankings for Apollo-9, a health tracker that didn’t exist. Inside hours, chatbots had been recommending the pretend health tracker and a few continued to take action a day after the fraud was uncovered. Liqing and different instruments are offered overtly on Chinese language e-commerce platforms like Taobao and JD.com, with pricing starting from roughly $520 to $4,765 for three-month subscriptions. One supplier informed Chinese language state media it had served greater than 200 purchasers throughout a number of industries, guaranteeing top-three placement on any AI platform.
With a purpose to higher perceive these developments and their influence, this text examines how the identical mechanism scales from business fraud to geopolitical disruption, utilizing the Russia-Ukraine conflict as a reside GEO lab. Leaked paperwork about Russia’s “Venture 2026” and Ukraine’s AI‑enabled counter‑operations present how affect campaigns are evolving from social feeds to the underlying sources that AI programs draw on — an angle largely absent from present data warfare debates.
Russia and the Nationwide Safety Case: Identical Playbook, Larger Stakes
When Russian operations exploit GEO of their conflict on Ukraine, they don’t seem to be simply spreading propaganda within the second; they’re attempting to grow to be the “floor reality” that AI programs summarize again to customers, analysts, journalists, and policymakers. In June 2026, Bloomberg reviewed 73 leaked paperwork from the Social Design Company (SDA), a sanctioned Moscow agency on the heart of Russia’s affect operations, describing a program its operators referred to as “Venture 2026:” a community of Wikipedia-style reference websites, media shops, and faux suppose tanks constructed to form what serps — and AI programs — deal with as dependable sources. The objective, as described within the leaked paperwork, is to “create another data ecosystem” by shaping not solely what folks see immediately but in addition what AI programs will later “know” about key leaders, the conflict’s origins, Ukraine’s conduct, NATO’s position, and Western help.
Russia’s GEO techniques construct on years of search engine marketing (website positioning) manipulation and are a part of a broadly reported sample of industrialized cognitive warfare that features deepfakes, cloned websites and different misleading content material. In 2023, a research printed within the Harvard Kennedy College Misinformation Overview examined pro-Kremlin makes an attempt to govern search engine outcomes and located that pseudo‑suppose tanks and propaganda shops corresponding to World Analysis and Strategic Tradition Basis had been amplified by way of backlink networks and low‑high quality websites. These shops had been best on conspiratorial searches involving, as an example, Ukraine President Zelensky, the place authoritative content material was sparse. Certainly, as famous by former CIA chief Jennifer Ewbank, using deepfakes to confuse, distort, or affect public opinion not solely happens in Ukraine however throughout Europe – all of which displays “the identical underlying actuality: the instruments for deception are sooner, cheaper, and extra accessible than the programs we depend on to detect or forestall them.” In different phrases, this isn’t nearly deception, however the erosion of belief itself.
Storm-1516, a documented Russian disinformation operation that has been lively since a minimum of 2023, has scaled sharply in 2026. Based on Bloomberg, the operation has produced greater than 190 false tales since 2023 that the outlet has been capable of determine. Primarily based on Bloomberg’s reporting, Meduza provides that within the first quarter of 2026 alone, Storm-1516 was producing pretend tales at twice the speed of the identical interval the earlier yr with, as an example, as of late March and early April 2026, supplies showing virtually each day. Meduza additionally experiences that greater than 40 p.c of Storm-1516’s fabrications have focused Ukraine, with one other third targeted on electoral processes in different international locations. Taken collectively, these experiences display that Storm-1516’s operations are finally aimed toward eroding Western help for Ukraine, swinging European elections and destabilizing NATO allies.
Taken along with the “Venture 2026” leaks, these findings counsel that Russia is now attacking each the content material layer (by way of artificial media) and the supply layer (by way of cloned Wikipedia‑fashion websites and faux suppose tanks) of knowledge ecosystems. Whereas artificial movies and tales are sometimes handled as quick‑time period deception, additionally they grow to be a part of the net document that future AI programs could ingest or retrieve, turning Russia’s layered affect structure into an extended‑time period GEO drawback, particularly in knowledge voids.
The Storm-1516 operation follows a clear sample. A pretend witness, typically an AI-generated video, seeds a believable however unverifiable story. Low-tier blogs and Telegram channels amplify it in a number of languages. Then much less rigorous Western shops choose it up, severing the hyperlink to the unique Russian operator. By the point the narrative reaches mainstream dialogue, the Russian fingerprint is gone. The brand new danger in immediately’s panorama is that the AI curation layer completes this laundering. It reads the now-repeated narrative throughout a number of sources and presents it as a impartial abstract.
Researchers on the Institute for Strategic Dialogue discovered that the chatbot DeepSeek was quoting VT International Coverage, an outlet identified to hold content material from Russian propaganda operations corresponding to Storm‑1516 and to have connections to the Kremlin‑linked Strategic Tradition Basis. U.S. and EU sources describe the Strategic Tradition Basis as an arm of Russian state pursuits.
A 2025 NewsGuard research additionally discovered ten of the main chatbots — together with ChatGPT, Claude, Gemini, and Copilot — collectively repeated false narratives from the pro-Kremlin Pravda community a couple of third of the time. It’s essential to notice {that a} 2025 research within the Misinformation Overview, responding on to the NewsGuard findings, discovered the quantity was nearer to five p.c and that these failures clustered in knowledge voids. Whereas the researchers discovered “little proof to help the grooming idea” and warned towards “the overhyped specter of Kremlin manipulation,” they acknowledged that knowledge voids “could also be artificially created” and that they may not dismiss the likelihood {that a} disinformation marketing campaign might goal them. Importantly, their audit got here fourteen months earlier than the leaked “Venture 2026” paperwork confirmed Russia explicitly concentrating on AI programs. The extra Russia makes an attempt to flood these knowledge voids, the extra possible it’s that future AI summaries about Ukraine will inherit its framing.
Easy methods to Construct Resilience – A Manner Ahead
Like its older cousin website positioning, GEO is inherently dual-use, but it surely sits in a regulatory vacuum as a result of it’s seen strictly as a client safety situation quite than a nationwide safety menace. The Apollo-9 experiment and the Storm-1516 operation show they’re two sides of the identical coin; the identical business techniques that manufactured demand for a non-existent health tracker can simply manufacture plausibility for distorted narratives a couple of geopolitical disaster corresponding to the continued Russia-Ukraine conflict. Historical past exhibits that with conventional serps, the market naturally incentivized tech corporations to deal with manipulation head-on as a result of mass spam threatened to destroy the consumer expertise for billions of individuals, straight endangering company enterprise fashions. Malicious overseas affect operations executing GEO are basically totally different as a result of they continue to be largely invisible to the mass market, with manipulation surgically clustered inside obscure knowledge voids, providing tech corporations no business incentive to self-police and leaving the knowledge terrain round a reside European conflict successfully undefended.
To bridge this hole, regulators can draw on the teachings of personal sector website positioning protection and public-private counter-disinformation efforts, however they need to understand that the very same playbook won’t work right here. Whereas personal builders can simply dismiss the Misinformation Overview research’s 5 p.c discovering as a suitable business error margin, this minor statistical anomaly sits exactly within the knowledge voids that Russia is actively attempting to colonize, and so it represents a main, unmonitored vector for overseas manipulation. The strategic danger is just not solely that GEO can mislead folks within the second, however that it could possibly reshape the evidentiary document on which establishments and AI programs will later rely; if hostile narratives are allowed to dominate lengthy‑tail matters and thinly documented episodes within the Russia-Ukraine conflict as an example, then future summaries, briefings, and even historic accounts danger being generated from polluted inputs.
The 2025 Misinformation Overview research recommends “warning banners for knowledge void queries” and elevated “audit entry,” however notes the banners are “utilized inconsistently” and the audit entry is “hindered by energy asymmetries.” As a result of the market won’t ever self-correct a menace it doesn’t financially register, defending the integrity of generative content material should transition from a voluntary company observe to a proper nationwide safety mandate.
Within the Russia-Ukraine conflict, these dynamics are already seen. Russian operations corresponding to Storm‑1516 use GEO‑fashion flooding and artificial witnesses to launder narratives in regards to the battle into the broader data atmosphere, whereas Ukrainian actors depend on AI‑enabled monitoring and evidentiary documentation to defend the integrity of the document. The competition is now not confined to what populations see on-line immediately; it’s over what AI programs will say is true in regards to the conflict tomorrow. Recognizing GEO as a nationwide safety drawback due to this fact modifications the governance query: the coaching, retrieval, and rating layers of generative programs at the moment are a part of the battlespace, and leaving this house unregulated is akin to leaving crucial data infrastructure undefended.
The Cipher Transient is dedicated to publishing a variety of views on nationwide safety points submitted by deeply skilled nationwide safety professionals. Opinions expressed are these of the writer and don’t symbolize the views or opinions of The Cipher Transient.
Have a perspective to share based mostly in your expertise within the nationwide safety discipline? Ship it to Editor@thecipherbrief.com for publication consideration.
Learn extra expert-driven nationwide safety insights, perspective and evaluation in The Cipher Transient
