Alphabet, the parent company of Google, is set to launch its artificial intelligence chatbot, Bard, in Europe and Brazil. This marks the largest expansion of the product since its initial release in the United States and the United Kingdom back in March. Bard, similar to Microsoft’s ChatGPT, utilizes generative AI to respond to queries in a manner that closely resembles human-like interactions.
The increasing capabilities of these AI chatbots have sparked a wave of concern among global tech leaders, leading some to advocate for a pause in their development. The potential impact of AI ranges from doomsday scenarios heralding the end of humanity to optimistic promises of solving climate change and other pressing issues.
In recent months, companies have poured billions of dollars into AI technology with hopes of reaping significant returns from advertising and cloud revenue. Mistral AI, a startup that is merely a month old, secured an impressive £86 million in seed funding to develop and train large language models. Additionally, entrepreneur Elon Musk recently announced the establishment of his AI startup called xAI, which boasts a team comprising former engineers from OpenAI and Google. Musk has been a vocal proponent of regulating the AI sector and has previously advocated for a temporary halt in AI development.
Meanwhile, Anthropic, an American AI firm, has released its own rival chatbot named Claude 2, which has the ability to summarize extensive blocks of text. Claude 2, publicly available in the US and the UK, employs a safety mechanism called “Constitutional AI,” which adheres to a set of principles guiding its text generation process.
However, the novelty appeal of AI chatbots might be dwindling, as recent web user statistics indicate a decline in monthly traffic and unique visitors to ChatGPT’s website in June for the first time.
Google has introduced new features to Bard, expanding its functionality worldwide. These enhancements include the chatbot’s capability to speak its responses aloud and its ability to process prompts accompanied by images. According to Jack Krawczyk, Google’s senior product director, Bard now supports collaboration in over 40 languages, enabling users to approach their ideas from different perspectives. Additionally, users can modify Bard’s tone and style of responses, choosing between simple, long, short, professional, or casual options. The updated version of Bard also allows users to pin or rename conversations, export code to various platforms, and utilize images in prompts.
The launch of Bard in the European Union faced delays due to concerns raised by the region’s primary data regulator, the Irish Data Protection Commission, regarding privacy safeguards. The commission deemed Alphabet’s explanations insufficient to justify a European launch, citing a lack of clarity on how the generative AI tool protects the privacy of European users. Alphabet has since engaged in discussions with the regulators to address transparency, choice, and control-related issues. Amar Subramanya, engineering vice president of Bard, stated that users could opt out of data collection but did not comment on the potential development of a dedicated Bard app, emphasizing that Bard is an experimental project driven by a commitment to boldness and responsibility.
In the United States, Google is facing a fresh class-action lawsuit alleging the unauthorized use of users’ personal information to train Bard. Eight individuals have filed the complaint in a federal court in San Francisco, seeking to represent millions of internet users and copyright holders. The plaintiffs argue that Google’s data scraping practices violated their privacy and property rights. Ryan Clarkson, the attorney representing the plaintiffs, emphasized that Google does not possess ownership rights over individuals’ personal information, creative works, or online expressions of identity.
The launch of Bard in Europe and Brazil and the growing concerns about the impact and regulation of AI continue to shape the landscape of artificial intelligence chatbots and their implications for privacy and user rights.
