Home / Technology / AI Firms Commit to Taming 'Hallucinations'
AI Firms Commit to Taming 'Hallucinations'
30 Apr
Summary
- Italy's antitrust authority closed investigations into three AI companies.
- Companies committed to better informing users about AI hallucination risks.
- DeepSeek will invest in technology to reduce inaccurate content generation.
Italy's antitrust authority, the AGCM, has closed investigations into three artificial intelligence companies. The companies, DeepSeek from China, Mistral AI from France, and Scaleup Yazilim Hizmetleri from Turkey, faced scrutiny over unfair commercial practices related to generative AI.
The primary concern for the AGCM was the potential for AI "hallucinations" – the generation of inaccurate or misleading information. To resolve these investigations, the three firms have provided binding commitments.
These commitments include improving user information regarding hallucination risks through website and app disclaimers. DeepSeek has also pledged to invest in technology aimed at reducing such inaccuracies, while acknowledging current limitations in preventing them entirely.
Furthermore, Scaleup's chatbot service, NOVA AI, will clarify its function. Consumers will be informed that it acts as a single interface to several chatbots rather than aggregating responses.
