ByteDance, the Chinese tech giant behind TikTok and Douyin, recently made headlines after firing an intern for “maliciously interfering” with the training of one of its AI models. This news, which initially spread through social media, has raised questions about the security and vulnerability of AI systems, even within leading tech companies, that AI is opening up new vectors of attack.
While ByteDance confirmed the intern’s dismissal, they were quick to deny the exaggerated claims of damage circulating online. The company clarified that the intern, who worked within the advertising technology team, had no prior experience with AI development. They also emphasised that the incident had no impact on their commercial online operations, including their popular Doubao chatbot, essentially dismissing the rumoured $10 million damage to their AI training system. However, AI opens up new vectors of attack, making it crucial to secure these systems.
This incident highlights a growing concern within the tech industry: the potential for internal sabotage or human error to disrupt AI development. As AI models become increasingly complex and integral to various applications, their security and integrity are paramount. Where there are new vectors of attack, more robust security is essential. And this isn’t a one-off; take a look at the global AI incident database to see what else has happened lately.
ByteDance’s response, including informing the intern’s university and industry bodies, underscores the seriousness of this issue. It also serves as a reminder to other companies about the importance of robust internal controls and security measures to protect their AI systems.
Key Takeaways:
- Insider threats are real: Even large tech companies with significant resources can be vulnerable to internal sabotage, especially in rapidly evolving fields like AI, which opens up new vectors of attack.
- Security is crucial: As AI becomes more integrated into various aspects of our lives, ensuring its security and integrity is critical
- Transparency is key: ByteDance’s proactive communication helped dispel rumours and manage the situation, highlighting the importance of transparency in such incidents.
This incident serves as a valuable lesson for the entire tech industry, emphasising the need for robust security measures and employee training to prevent similar incidents in the future.
For non tech businesses, the lesson is harsh – you can’t rely on your pre-AI cyber security measures to be good enough for the new world we’re creating.
