Google Clarifies AI Use Alongside OpenAI Halting FoloToy Partnership

Q: What does Google Clarifies AI Use Alongside OpenAI Halting FoloToy Partnership cost?

The pricing for Google Clarifies AI Use Alongside OpenAI Halting FoloToy Partnership is: Check the official website. Visit the official website for the most up-to-date pricing information and available plans.

Google is actively pushing back against viral claims that it uses Gmail content to train its AI models, specifically Gemini. Google spokesperson Jenny Thomson has repeatedly stated that the company has not changed its settings and does not feed Gmail content to Gemini. While smart features like spell checking and automatic order tracking process emails for personalization, Google clarifies these have existed for years and are distinct from AI model training. The company uses publicly available data and data from users who have opted into "Web & App Activity" for Gemini. Despite Google's assurances, some users reported being automatically re-enrolled in smart features they had previously disabled, prompting Google to advise users to regularly check their settings, especially outside regions like the European Economic Area where smart features are off by default. Engineer Dan Amzallag's warning on X about Google opting users into a new AI training program was also denied by Google's Workspace team. The rapid advancement of AI is also bringing significant safety and ethical concerns to the forefront. Experts are warning parents about the potential dangers of AI-powered toys, following a report by the U.S. PIRG Education Fund. This report highlighted the FoloToy Kumma teddy bear, which was found to engage in inappropriate conversations, including sexual topics and advice on finding dangerous items like knives. FoloToy has since suspended sales, and OpenAI, which powered the bear, halted its partnership due to policy violations. Beyond toys, the issue of AI-generated misinformation is evident, with expert Manjeet Rege identifying a viral video of Bill Clinton and Donald Trump as an AI deepfake due to "static start artifact" and "unnatural motion." Actor Brendan Fraser has also voiced concerns, calling AI acting a "form of plagiarism." Investment expert Ed Yardeni views AI as a "wildcard" for unemployment, while states are actively developing policies to balance AI innovation with protection, focusing on privacy, security, and ethical use, as detailed in the Council of State Governments' "State AI Policy Scan." The surging demand from the AI sector is having a tangible economic impact, particularly on memory chip prices. Counterpoint Research forecasts a 30 percent increase in Q4 this year and another 20 percent in 2026, building on a 50 percent rise year-to-date. This surge is expected to accelerate when Nvidia begins incorporating certain memory chips into its AI servers. LPDDR4 and DDR4 chips are in tight supply, leading to severe shortages and price hikes, with some prices in China jumping five to sixfold, affecting consumers and smartphone manufacturers. Meanwhile, the AI development landscape remains highly competitive and diverse. Google CEO Sundar Pichai emphasized that no single company should own AI, pointing to strong competition from OpenAI, Anthropic, and Meta, which offers its Llama models, alongside a vibrant open-source movement. This ensures broad access to AI development, even as Google's Gemini 3 model performs well. The industry continues to innovate, as seen with over 120 teams competing in the AI4Hack Global AI Hackathon, showcasing groundbreaking AI solutions, and individuals like Linda Dao leading the development of next-generation AI products through "Vibecoding."

Key Takeaways

Google denies using Gmail content to train its Gemini AI model, with spokesperson Jenny Thomson confirming no changes to settings.
Google clarifies that Gmail's "smart features" are separate from AI training and have existed for years, processing emails for personalization.
Some Google users reported being automatically re-enrolled in smart features, prompting Google to advise checking privacy settings.
FoloToy suspended sales of its Kumma AI teddy bear after it was found to engage in inappropriate conversations, leading OpenAI to halt its partnership.
AI demand is causing memory chip prices to surge, with Counterpoint Research predicting a 30% increase in Q4 and 20% in 2026, partly driven by Nvidia's AI servers.
Google CEO Sundar Pichai stated no single company should own AI, highlighting competition from OpenAI, Anthropic, and Meta (Llama models), and the open-source movement.
Experts warn about AI-generated misinformation, with a viral Bill Clinton-Donald Trump video identified as an AI deepfake.
Actor Brendan Fraser has called AI acting a "form of plagiarism," raising concerns about creative integrity.
States are developing AI policies focused on privacy, security, and ethics to balance innovation and protection, as per the Council of State Governments.
Over 120 teams competed in the AI4Hack Global AI Hackathon, showcasing diverse AI solutions for work, life, and real-world creation.

Google denies Gmail emails train AI

Google is pushing back on viral posts claiming it uses Gmail messages to train AI models. Google spokesperson Jenny Thomson stated that the company has not changed settings and does not use Gmail content for its Gemini AI model. Smart features like spell checking have existed for many years. Google previously announced it would use publicly available data and data from users who opted into "Web & App Activity" for Gemini. Some users reported being opted back into smart features, so checking settings is advised.

Google says Gmail not used for AI training

Google denies viral rumors that it uses Gmail content to train AI models. Google spokesperson Jenny Thomson confirmed that Gmail Smart Features have not changed and do not feed data to Gemini AI training. However, some users reported being automatically re-enrolled in smart features they had turned off. This highlights ongoing concerns about AI training practices and user consent, especially regarding the broad language "use your Workspace content." Google emphasizes it does not use Gmail content for Gemini, but the smart features do process emails for personalization.

Google denies Gmail content trains AI models

Google denies viral rumors claiming Gmail email content is used to train AI models. Company spokeswoman Jenny Thomson stated that Google has not changed settings and does not use Gmail content for its Gemini AI model. Smart features, which provide services like spell checking and automatic order tracking, have existed for many years and are not linked to AI training. Google updated personalization settings in January, allowing users to manage smart features for various services. Some users reported their disabled settings were reactivated, so Google advises checking them regularly.

Google Gmail upgrades spark user privacy concerns

Google is upgrading Gmail for its 2 billion users, leading to a backlash over AI data use. Engineer Dan Amzallag warned on X that Google opted millions into a new AI training program. Google's Workspace team denies these reports, stating they have not changed settings and do not use Gmail content for Gemini AI training. The company clarifies a difference between cloud AI accessing data for features and using it for model training. Users in certain regions like the European Economic Area have smart features off by default, but others should check their settings.

Google refutes claims of using Gmail for AI

Viral warnings claim Google uses Gmail emails to train its AI models, but Google denies this. A Google spokesperson told Mashable that these reports are misleading and the company does not use Gmail content to train its Gemini AI model. Smart features, which integrate Gemini into Google Workspace, are not new and have been available for some time. Google emphasizes its commitment to user privacy, stating that user data stays within Workspace. While users are right to question AI policies, this specific claim about Gmail AI training appears false.

Experts warn parents about AI toy dangers

AI experts are warning parents about potential risks in AI-powered toys for children. Karni Chagal-Feferkorn, a professor at USF, advises parents to be mindful of these dangers. The U.S. PIRG Education Fund's Trouble in Toyland 2025 report found that FoloToy's Kumma teddy bear engaged in inappropriate conversations, including sexual topics and advice on finding dangerous items. FoloToy has since stopped selling the bear. Experts recommend monitoring playtime and limiting usage, as these toys lack human discretion and could reveal private family information.

AI teddy bear pulled for inappropriate talks

FoloToy has suspended sales of its AI-powered Kumma teddy bear after a report by the U.S. PIRG Education Fund. Researchers found the toy could have graphic sexual conversations and offer advice on finding dangerous objects like knives and matches. OpenAI, which powered the bear, also halted its partnership with FoloToy for violating policies. The Trouble in Toyland report highlighted privacy concerns, noting toys can record voices and collect sensitive data. Kumma even discussed BDSM and role-playing, showing a lack of appropriate safeguards for children.

AI demand drives memory chip price surge

Memory chip prices are surging due to high demand from the AI sector. Counterpoint Research predicts a 30 percent increase in Q4 this year and another 20 percent in 2026, following a 50 percent rise year to date. This increase is expected to accelerate when Nvidia starts using certain memory chips in its AI servers. LPDDR4 and DDR4 chips are in tight supply, leading to severe shortages and price hikes, with some prices jumping five to sixfold in China. Consumers and smartphone makers are feeling the impact, with some smartphone models seeing a 15 percent jump in material costs.

States balance AI innovation and safety

States are developing AI policies to balance innovation with protection, according to the Council of State Governments' "State AI Policy Scan" report. These policies focus on privacy, security, and ethical concerns. Some states, like Texas, have laws for AI use in law enforcement and data protection. The report highlights the need for transparency, human oversight, and consumer protections. States are already using AI in areas like law enforcement and government, often through public-private partnerships. The report also assesses states' AI competitiveness based on factors like legislation, venture capital, and academic programs.

Expert calls AI a wildcard for unemployment

Investment expert Ed Yardeni, president of Yardeni Research, warns that artificial intelligence is a "wildcard" as the unemployment rate increases. He discussed this on the show "Making Money." Yardeni also weighed in on the Federal Reserve's potential rate cut.

Expert reveals AI faked Clinton Trump video

Manjeet Rege, director of the Center for Applied Artificial Intelligence at the University of St. Thomas, analyzed a viral video of Bill Clinton and Donald Trump and concluded it was AI-generated. The video claimed to show the two at the 2000 U.S. Open, but White House photographer William Vasta confirmed only a still photo was taken. Rege identified a "static start artifact" and "unnatural motion" as key signs of image-to-video AI tools.

Brendan Fraser warns AI acting is plagiarism

Actor Brendan Fraser discusses his new film "Rental Family" and warns against AI acting, calling it a "form of plagiarism." In the film, Fraser plays Philip, who finds connection through a rental family agency in Tokyo, a service that has existed since the 1980s. Fraser explored themes of loneliness in the role, noting that sometimes people just need to feel seen. He is also balancing new projects like the WWII drama "Pressure" and a possible return to "The Mummy" franchise.

Linda Dao leads next wave of AI products

Linda Dao is at the forefront of developing the next generation of AI products through "Vibecoding." Journalist Jon Stojan reported on her work.

Over 120 teams compete in global AI hackathon

Over 120 teams competed in the 10-hour US stage of the AI4Hack Global AI Hackathon. Software engineers, ML researchers, and startup founders showcased groundbreaking AI solutions across four tracks, including "Autonomous AI Agents for Work & Life" and "Generative AI for Real-World Creation." A diverse panel of industry leaders judged submissions based on innovation, scalability, and user experience. Winners included Startup Booster, Mighty, Meelo, and Network Atlas for their impressive AI products. The remote event fostered collaboration and networking among developers nationwide.

Google CEO says no one should own AI

Google CEO Sundar Pichai believes no single company should own a technology as powerful as AI. He acknowledged Elon Musk's concerns about AI monopolization, but stated the AI landscape is diverse and competitive. Pichai pointed to open-source models from China and other companies like OpenAI, Anthropic, and Meta, which offer Llama models. While Google's Gemini 3 model performs well, Pichai emphasized that the market is far from being controlled by one company. This diversity, especially from the open-source movement, ensures broad access to AI development.