In the Age of Artificial Intelligence, Disinformation Targets Both Human and Machine Perception
Autonomous disinformation and artificial intelligence swarms evolving in the age of intelligence create new dangers targeting both the human mind and the machine perception of artificial intelligence systems.
Disinformation activities in the age of artificial intelligence threaten the global information ecosystem by targeting both humans and machines through autonomous AI agents and AI swarms.
Transformation of the Global Information Ecosystem
Technological developments have made the global information ecosystem more digital, complex, and fragmented. Artificial intelligence sits at the center of this transformation, supporting data-driven decision-making.
Generative AI models have flooded the information ecosystem with content of inconsistent quality since 2022, and it is estimated that approximately 50 percent of newly published articles are AI-generated.
Cognitive Warfare and Disinformation
While disinformation is intentionally spread for manipulation purposes, the digital era has intensified the scope and reach of these activities.
Companies are at risk due to coordinated disinformation in supply chains and stock-sensitive news, or deepfake content impersonating executives.
Autonomous Agents and Artificial Intelligence Swarms
An estimated 28.6 million agents were in use across businesses in 2025, and this number is projected to exceed 2 billion by 2030.
Custom AI agents operating in coordinated networks can greatly amplify disinformation campaigns with increased autonomy and speed.
Targeting Machine Perception
Cognitive warfare can target not only humans, but also the machine cognition of artificial systems that carry out processes similar to human perception.
With the proliferation of synthetic data, risks of data poisoning are increasing, and misleading content can infiltrate the training data of large language models.
Building Cognitive Security
Protecting the information ecosystem requires a multi-layered approach that will increase cognitive security and resilience for both humans and machines.
This process encompasses educational and political measures for humans, and reliable datasets and cross-platform monitoring systems for machines.