Mandiant: the use of AI to conduct politically-motivated online influence campaigns has grown in recent years, but the impact of such campaigns has been limited
Google-owned U.S. cybersecurity firm Mandiant said on Thursday it had seen increasing use of artificial intelligence (AI) …
Context & Ripple Effects
Mandiant’s assessment provides an early baseline in a coverage arc where AI’s political use was already prompting calls for guardrails around AI-generated campaign advertising. Its key distinction is between growing adoption of a tool and demonstrated influence on audiences.
Later coverage reinforces that distinction: OpenAI reported disrupting five covert influence operations, while a DeepMind review found political deepfakes were the most common observed AI misuse. Detection and prevalence do not, by themselves, establish persuasive impact.
First-order effects
- Mandiant’s finding gives defenders and policymakers a more calibrated threat picture: AI-assisted influence activity is increasing, but the reported evidence does not show commensurate effectiveness.
- The immediate focus shifts toward identifying coordinated influence operations and measuring their reach or behavioral effects, rather than treating AI-generated content alone as proof of successful manipulation.
Second-order effects
- Platforms and AI providers face pressure to maintain disruption and monitoring capabilities for political misuse even when observed campaigns have limited impact.
- Election-related guardrail debates become more dependent on evidence of distribution, targeting, and audience response—not just on the availability of generative tools.
Third-order effects
- If AI lowers the cost of producing influence content faster than it improves persuasion, the durable challenge will be measuring campaign outcomes and coordinating responses across model providers, platforms, and security firms.
- The pattern points toward dual-use AI governance centered on misuse monitoring and provenance, with the scope of intervention likely to depend on whether effectiveness catches up to production scale.
The trend: Political AI risk is moving from concern over synthetic-content volume toward evidence-based assessment of whether coordinated campaigns can actually change public outcomes.