Google Bard: AI chatbot now available in Europe
Supports more than 40 languages
Google parent Alphabet is launching its AI-driven chatbot, Bard, across Europe and several other countries, significantly increasing its reach.
"Today we're announcing Bard's biggest expansion to date," wrote Bard's product lead Jack Krawczyk and vice president Amarnag Subramanya.
"It's now available in most of the world, and in the most widely spoken languages. And we're launching new features to help you better customise your experience, boost your creativity and get more done."
After Bard's initial introduction in February, Google encountered delays in launching the chatbot in the European Union.
The Irish Data Protection Commission (DPC) raised privacy concerns, holding up the roll-out.
The DPC said Google had not supplied enough information about its generative AI tool's privacy safeguards to warrant a launch within the EU.
Ireland serves as the European headquarters for many prominent US tech companies, including Google.
Google has confirmed that it has engaged with European watchdogs to address issues they raised surrounding transparency, choice and user control.
Graham Doyle, spokesperson and deputy commissioner of the DPC, acknowledged that Google has implemented several modifications in Bard prior to its launch in Europe, including enhanced transparency measures and revised user controls.
"Google have agreed to carrying out a review and providing a report to the DPC after three months of Bard becoming operational in the EU," Doyle added.
"We will be continuing our engagement with Google in relation to Bard post-launch."
At a briefing with journalists, Amar Subramanya said users have the option to opt out of data collection.
Bard goes global
As well as the EU, Google also launched Bard in Brazil and other countries on Thursday.
Google has significantly expanded the chatbot's language capabilities. It now supports more than 40 languages, including Arabic, Chinese, German, Hindi and Spanish. Previously, it was only available in three languages: English, Japanese, and Korean.
Google is also introducing new features, including the ability to receive audio responses from Bard and choose from five different response styles: simple, long, short, professional or casual.
Users have the option to pin or rename conversations, export code to various platforms and incorporate images in prompts.
Additionally, Google is integrating the capabilities of Google Lens into Bard. Users can now upload images with prompts, and Bard will analyse the images to provide information.
Subramanya did, however, acknowledge that there are instances where the model may not provide accurate responses when faced with complex questions.
"It can also reflect biases from the training data. It could be adversarial to the training model. These are some of the limitations of Bard. We do use answers from search, to keep it up to date," he said.
Many companies have invested heavily in AI over the last six months.
The latest figure to enter the game is Elon Musk, who this week announced a new AI start-up called xAI.
The team at xAI comprises several engineers who have previous experience working at OpenAI and Google (obviously an entirely different situation to Meta hiring former Twitter engineers to work on Threads, for which Musk threatened to sue - Ed.).
Musk had previously advocated for pausing AI developments until regulatory measures could be introduced.
Last month, the European Parliament reached an agreement to finalise the AI Act, marking a significant step towards enacting the world's first legislation specifically targeting AI.
The proposed bill encompasses provisions addressing generative AI systems like ChatGPT and Dall-E, which have the capability to generate text, images, and other forms of media.
The Act outlines a categorisation framework for AI solutions based on the potential risks they pose, dividing them into four categories: minimal, limited, high, and unacceptable.
The assigned category determines the specific governance rules and regulations that apply to each solution.
The AI Act is projected to become law by 2025, but before that, it must undergo voting processes by the European Commission, Parliament, and Council.