Google plans to reintroduce the Gemini AI picture generator in a ‘few weeks’ in response to increasing criticism of its inaccurate images.

After withdrawing its artificial intelligence image generation tool on Thursday due to a series of controversies, Google intends to reintroduce the product soon, according to Demis Hassabis, CEO of Google DeepMind. The image generator was initially launched this month through Gemini, Google’s primary suite of AI models, enabling users to input prompts for image creation. Over the past week, widespread social media circulation highlighted historical inaccuracies and questionable responses, prompting the tool’s temporary removal.

Hassabis addressed the situation during a panel at the Mobile World Congress conference in Barcelona on Monday, stating, “We have taken the feature offline while we fix that. We are hoping to have that back online very shortly in the next couple of weeks, few weeks.” He acknowledged that the product was not functioning as intended.

Alphabet shares experienced a 4.4% decline on Monday, closing at $137.57.

The controversy arose following Google’s recent high-profile rebrand, which included renaming its chatbot from Bard to Gemini and introducing a new app with subscription options. Users reported issues such as historical inaccuracies and diverse representations when requesting images through the Gemini image generator, leading to concerns about the tool’s functionality and ethics.

Examples of problematic requests included asking for a German soldier in 1943, which resulted in a racially diverse set of soldiers, and a request for a historically accurate depiction of a medieval British king, which generated racially diverse images, including a woman ruler. Similar outcomes occurred with requests for images of U.S. founding fathers, an 18th-century king of France, a German couple in the 1800s, and even Google’s own founders.

The incident has reignited debates within the AI industry, with some labeling Gemini as too “woke” or left-leaning, while others criticize Google’s approach to AI ethics. Sundar Pichai, Alphabet CEO, has faced criticism for the company’s handling of AI projects, including the botched rollout of Bard in the past.

The controversy extends beyond Gemini’s image generator, as a text-based query asking the Gemini chatbot about the negative impact of Adolf Hitler versus Elon Musk’s tweeting went viral. Google acknowledged the issues with Gemini’s image generation and announced a pause on generating images of people, with plans to release an improved version soon.

Google is heavily investing in AI agents, aiming to expand their capabilities from tasks like summarizing and generating to-do lists to more complex functions such as scheduling, travel booking, and job-specific tasks. The recent changes to Gemini are seen as a step toward building a comprehensive AI assistant, according to Sissie Hsiao, a vice president at Google.