“AI is neither good, nor bad, nor is it neutral.” Invoking Kranzberg’s First Law, Alexios Mantzarlis (Cornell Tech / Indicator Media) opened his keynote at Media Party New York 2026 by dismantling the binary debate around artificial intelligence. He presented a pragmatic view of a tool that, while threatening digital safety, offers journalists unprecedented investigative scale.
Mantzarlis addressed the “reckless deployment” of AI—from predatory “slop” siphoning newsroom revenue to the rise of synthetic abuse and “disinfobate.” His message was clear: the future of media depends on mastering the middle ground, where human editorial judgment directs the raw power of the machine to protect information integrity.
The Architecture of a Digital Disaster
Drawing from his frontline experience at Indicator Media, Mantzarlis detailed how AI has accelerated the degradation of the information ecosystem. He moved beyond theoretical risks to identify specific, predatory models that are currently targeting both journalists and the public:
- Image-Based Sexual Abuse: Citing UNICEF data, he highlighted the “industry” behind non-consensual synthetic nudes. He noted that the digital safety of young women is increasingly compromised by the rapid expansion of these tools and deficient platform oversight.
- “Disinfobate” and Scam Compounds: Mantzarlis described a sophisticated pipeline where AI-driven accounts farm engagement—often using AI avatars of real media figures—to reach monetization thresholds. Once an audience is built, it is sold to scam compounds that target the “digitally naive” with financial fraud.
- The Rise of “Predatory Slop”: He exposed the emergence of automated podcast networks that use AI to instantly scrape and summarize original reporting from established news organizations and student publications. These networks often publish audio versions of breaking stories within minutes of the original article’s release, siphoning away traffic and revenue.
Scaling Accountability: AI as a Force Multiplier
Despite these threats, Mantzarlis argued that total rejection of AI is a luxury journalists can no longer afford. Balancing his reporting with his leadership role at Cornell Tech, he demonstrated how AI agents act as a force multiplier for complex investigations that would otherwise require massive resources.
I can and must use AI responsibly to scale up my own impact.”
He shared a case study where he utilized AI agents to identify and audit a network of TikTok scams. By using the technology to expand his initial data set, he was able to present the platform with evidence of 960 scam videos instead of just a handful, forcing a much more comprehensive takedown.
Furthermore, he applied AI to analyze 2.4 million Community Notes on X. The agent allowed him to cluster the data and identify a specific “no-note” gaming strategy used to suppress fact-checks on sensitive topics during the UK election—an audit that would have been manually impossible for such a small team.
A Manifesto for the Hybrid Journalist
To navigate this landscape without losing editorial spirit, Mantzarlis is codifying a four-pillar policy for AI integration:
- Never Delegate Thinking: AI is a tool for execution, not strategy. The initial hunch, the investigation’s direction, and the final narrative must remain strictly human.
- Proportional Disclosure: Transparency should be as granular as the technology’s involvement. If AI significantly shaped the data analysis or visualization, the audience deserves a clear explanation of its role.
- Data-First Verification: Journalists must maintain access to the raw data the AI is processing. “Verify everything” remains the golden rule; if you cannot see the source data, you cannot trust the summary.
- Resilient Tooling: Recalling the sudden ending of previous industry-standard tools, Mantzarlis warned against creating single points of failure. Journalists must build “fallbacks” and avoid over-relying on any single proprietary platform that can change its access or terms of service overnight.
The Human Verdict
Mantzarlis concluded with a call to move past the dichotomy of AI alarmists versus enthusiasts. The future of journalism, he suggested, lies in the hands of those who can fight the “two wolves” simultaneously: acknowledging that AI is a disaster for information quality while using it to hold the very systems that created it accountable.
At Media Party, the sentiment was finalized: the genie is out of the bottle. Our task is to ensure that while the machine provides the scale, the human provides the purpose.
Stay tuned for more updates and deep dives from Media Party New York 2026!
