ByteDance Fires Intern for 'Malicious Interference' in AI Training, Denies $10M Damage Claims

October 22, 2024
ByteDance Fires Intern for 'Malicious Interference' in AI Training, Denies $10M Damage Claims
  • As a leader in algorithm development, ByteDance employs over 110,000 people globally and continues to be a major player in the social media industry, driven by the popularity of TikTok and Douyin.

  • ByteDance recently terminated an intern for allegedly 'maliciously interfering' with the training of its AI models, an incident that occurred in August 2024.

  • The intern, who was part of the commercial technology team, was accused of serious disciplinary violations and is now the subject of reports to their university and relevant industry bodies.

  • The alleged sabotage involved the intern planting malicious code that slowed down TikTok's AI model training, which relies on over 8,000 NVIDIA H100 GPUs.

  • This incident gained significant attention on social media, particularly concerning ByteDance's popular generative AI model, Doubao.

  • The company emphasized that its formal commercial projects and large language AI models were not affected by the intern's actions.

  • This incident occurs amid a broader context of increasing scrutiny on tech companies regarding the safety of generative AI models and the impact of social media.

  • Despite ByteDance's efforts to clarify the situation, skepticism persists online, with some commenters questioning the company's statements about the intern's team affiliation.

  • ByteDance has denied claims that the incident resulted in over $10 million in damages, asserting that the disruption affected only a research project and did not impact commercial operations.

  • As ByteDance continues to develop generative AI technologies, including the Doubao chatbot, it faces scrutiny regarding the safety and ethical implications of its AI models.

  • US officials have raised concerns about ByteDance's potential threat to national security, especially as the company reportedly develops new AI models using Huawei chips.

  • Additionally, ByteDance's AI algorithms are under scrutiny for their potential negative effects on children, as revealed in leaked internal documents.

Summary based on 18 sources


Get a daily email with more Tech stories

Sources





More Stories