Irish Watchdog Probes Google’s AI Data Practices

The Irish Data Protection Commission (DPC) has initiated a “cross-border statutory inquiry” into Google’s advanced artificial intelligence (AI) model to investigate whether the tech giant has complied with European data protection regulations in handling users’ personal data. This move comes as part of the DPC’s mandate to ensure companies operating in the European Union (EU) adhere to stringent privacy standards.

According to The Hacker News, the focus of the inquiry is to determine if Google conducted a required Data Protection Impact Assessment (DPIA) before processing personal data linked to European Economic Area (EEA) and EU citizens. This assessment is mandated by Article 35[2] of the General Data Protection Regulation (GDPR) and is crucial when dealing with high-risk data processing activities, such as those involved in the development of AI systems.

At the center of the investigation is Google’s Pathways Language Model 2 (PaLM 2), a sophisticated AI language model known for its enhanced multilingual, reasoning, and coding capabilities. PaLM 2 was unveiled in May 2023 and represents Google’s latest advancement in the AI field. With Google’s European headquarters in Dublin, the DPC holds the responsibility of ensuring that the company complies with the EU’s strict data privacy rules.

Related: US Tightens Grip on AI: New Reporting Rules for Developers and Cloud Providers

Per The Hacker News, the DPC emphasized that the inquiry is essential to safeguarding individuals’ fundamental rights and freedoms, especially as AI systems that process personal data pose significant privacy risks. The DPC’s investigation will assess whether the proper protocols were followed in Google’s handling of such sensitive data.

This scrutiny comes at a time when the use of personal data for AI development is under increasing regulatory attention. Social media platform X (formerly Twitter) recently agreed not to use personal data from European users to train its AI chatbot, Grok, without obtaining explicit consent. Similarly, Meta (Facebook’s parent company) faced criticism for collecting public data from Australian users for training its Llama AI models. Following intervention by the DPC, Meta has paused its AI-related activities in Europe and suspended the use of generative AI in Brazil after facing objections from the country’s data protection authority.

The AI sector has faced growing global scrutiny over its data handling practices. In 2023, Italy’s data protection authority temporarily banned OpenAI’s ChatGPT, citing violations of local data privacy laws. These actions underscore the challenges that tech companies face in balancing AI innovation with privacy regulations across different regions.

Source: The Hacker News