Hackers Hate AI Slop Even More Than You Do

Why It Matters

The backlash highlights a cultural friction point that could shape how AI tools are adopted—or resisted—within illicit online ecosystems, influencing future threat‑actor tactics and security defenses.

Key Takeaways

• Study examined 97,895 AI mentions on cybercrime forums (2022‑2023)
• Low‑level hackers resent AI‑generated posts, citing loss of credibility
• AI hasn't lowered entry barriers for basic cybercrime activities
• Sophisticated actors exploit AI loopholes, but remain cautious of guardrails
• Some forum users consider AI assistants for grammar, not full automation

Pulse Analysis

The rise of generative AI has reshaped many digital domains, but its penetration into underground cybercrime forums is meeting unexpected resistance. Researchers from the University of Edinburgh, Cambridge and Strathclyde analyzed almost 100,000 AI‑related conversations spanning from ChatGPT’s debut in 2022 through 2023. Their findings reveal a clear split: elite threat actors are already leveraging AI to craft malicious code, automate phishing, and bypass model guardrails, while the broader community of low‑skill hackers balks at AI‑filled posts that they view as lazy, low‑quality content. This cultural pushback underscores that reputation remains a core currency in these illicit markets, and automated output threatens the perceived expertise of participants.

For the cybersecurity industry, the mixed reception of AI on illicit platforms signals both risk and opportunity. On one hand, the lack of a dramatic skill‑level reduction suggests that AI alone will not democratize sophisticated attacks, buying time for defenders to develop counter‑measures. On the other hand, the growing chatter about AI‑enhanced marketplaces and tools indicates that threat actors are experimenting with ways to streamline data sales and automate routine fraud, such as SEO manipulation and romance scams. Security firms like Flashpoint are already tracking discussions around models like Anthropic’s Claude Mythos Preview, highlighting the need for continuous monitoring of AI‑related threat intel.

Looking ahead, the tension between human‑centric community norms and the efficiency gains offered by AI may shape the evolution of cybercrime forums. Some users are tentatively open to AI assistants that improve post grammar without fully automating content, hinting at a hybrid future where AI augments rather than replaces human interaction. However, vocal opposition—evident in threads demanding an end to "AI shit"—could lead to stricter moderation or the emergence of AI‑free sub‑communities. For defenders, understanding these dynamics is crucial: it informs threat modeling, helps anticipate which attack vectors may gain traction, and guides the development of proactive defenses against AI‑enabled fraud.