AI chatbots are spreading voter misinformation as midterm elections approach. The Washington Post tested ChatGPT by asking it to create voter guides for Missouri and Wyoming, and the chatbot provided false information such as listing a Senate seat that was not up for re-election. Out of 3,000 AI-generated voting guides, many contained wrong voting details. Chatbots failed to identify the correct congressional district 48 percent of the time. Despite these errors, some voters are increasingly turning to AI chatbots instead of the internet or friends for election research. The Brennan Center for Justice studied this trend and explored what happens when voters rely on AI tools for election information.
Anthropic has restricted its Claude AI software from cloning video games. An update to Claude now blocks the creation of exact copies of existing games, allowing only modified versions. User chasm was among the first to notice the change on Twitter. However, experts say the new rules will not slow the growing wave of AI-assisted game copying. They note the technology is improving and becoming harder to detect, leaving industry creators concerned about the ongoing impact on their work.
Chinese robot makers are shifting focus to embodied AI software that helps robots understand tasks. Funding for Chinese embodied AI companies reached 93.5 billion yuan across 322 deals in the first half of 2026, five times the amount from a year earlier. Meanwhile, researchers warn that AI is slowly eroding human control through a risk called gradual disempowerment. Jay Chooi of Robocurve says organizations have incentives to phase out humans. Jan Kulveit warns governments may become less democratic, and Margaret Mitchell of Hugging Face says people may lose skills and struggle to fix problems when they occur.
Requiring AI use at work can backfire. Some workplaces hold mandatory AI training sessions or require employees to use a specific number of tokens or log set hours on AI platforms. At the Nonprofit Leadership Conference, Jack Clark of Anthropic spoke about AI's accelerating pace. Tiff Macklem of the Bank of Canada says AI will displace workers in certain professions, while Kristalina Georgieva of the IMF notes that service jobs pay little and new job creation is hard to predict. Ayesha Kahnna of Addo AI highlights Singapore's subsidized skills training programs as a model for adaptation.
Small businesses are leading the AI economy. A Fast Company survey of Goldman Sachs 10,000 Small Businesses alumni shows 86% now use AI, though only 15% have fully integrated it into their operations. Pompey compared early AI adoption to blacksmiths who adapted to automobiles. Google also launched a free learning platform called Skilling Africa with Google for Nigerians, offering over 15 digital and AI courses, some offline. The platform aims to train three million Africans by 2030 and includes an AI study assistant called SCOUT. Separately, researchers warn that advances in AI voice cloning make it increasingly difficult to tell real recordings from AI-generated ones, as cloned voices now include fake breaths and pauses that mimic natural speech.
Key Takeaways
- Chatbots failed to identify the correct congressional district 48 percent of the time when generating voter guides, spreading misinformation ahead of midterm elections.
- Anthropic restricted Claude from creating exact copies of existing video games, but experts say the new rules will not stop the AI copying boom.
- Funding for Chinese embodied AI companies reached 93.5 billion yuan in the first half of 2026, five times the amount from a year earlier.
- Researchers warn of gradual disempowerment as AI takes over more tasks, with Hugging Face's Margaret Mitchell warning people may lose skills they once used.
- Requiring AI use at work can backfire, with mandatory training sessions and token requirements producing poor results.
- Jack Clark of Anthropic spoke at the Nonprofit Leadership Conference about AI's accelerating pace and impact on global jobs.
- 86% of small businesses now use AI, but only 15% have fully integrated it into their operations.
- Google launched Skilling Africa with Google, a free platform aiming to train three million Africans by 2030 with AI and digital courses.
- AI voice cloning has advanced to the point where cloned voices include fake breaths and pauses, making detection difficult with no single reliable tool.
- The Brennan Center for Justice studied the trend of voters using AI chatbots for election research instead of traditional sources.
ChatGPT spreads voter misinformation as midterm elections approach
The Washington Post tested ChatGPT by asking it to create voter guides for Missouri and Wyoming. The chatbot gave false information like listing a Senate seat that was not up for re-election. Out of 3,000 AI-generated voting guides, many contained wrong voting details. Chatbots failed to identify the correct congressional district 48 percent of the time. Few voters appear to be using these chatbots for election information.
Voters increasingly use AI chatbots for election research
Some voters are turning to AI chatbots instead of the internet or friends to research candidates and political issues. The Brennan Center for Justice studied this trend. Abdiaziz Ahmed, a technology policy strategist for the center, discussed the findings on Minnesota Now. The conversation explored what happens when voters rely on AI tools for election information.
Voters use AI chatbots for Election Day guidance
More voters are using AI chatbots to find election-related information. NBC News' Savannah Sellers interviewed a voter who used AI to navigate their ballot. The segment examined the risks of depending on chatbots for voting information on Election Day.
Anthropic restricts Claude from cloning video games
An update to Anthropic's Claude AI software appears to block the creation of exact copies of existing video games. Claude can now only create modified versions rather than outright clones. This change comes as controversy grows around AI-created copies of video games.
Claude's new anti-cloning rules will not stop AI copying boom
Anthropic added a new system prompt to stop Claude from recreating specific existing video games. User chasm was among the first to notice the change on Twitter. Experts say the new rules will not slow the growing wave of AI-assisted game copying. They say the technology is getting better and harder to detect. Industry creators worry about the impact on their work as this remains a ongoing battle.
China robot makers shift focus to AI brains
Chinese robot companies have built machines that can run and dance. Now they are focusing on embodied AI software that helps robots understand tasks. Funding for Chinese embodied AI companies reached 93.5 billion yuan across 322 deals in the first half of 2026. This is five times the amount from a year earlier.
Researchers warn AI is slowly eroding human control
Researchers describe a risk called gradual disempowerment as AI takes over more tasks. Jay Chooi of Robocurve says organizations have incentives to phase out humans. Jan Kulveit warns governments may become less democratic if they no longer need citizens. Margaret Mitchell of Hugging Face says people may lose skills they once used and struggle to fix problems when they occur.
Requiring AI use at work can backfire
AI literacy is becoming a required skill for many jobs. Some workplaces hold mandatory AI training sessions. Others require employees to use a specific number of tokens or log set hours on AI platforms. The article explores why forcing workers to use AI can produce poor results.
Panel discusses AI's impact on global jobs
Jack Clark of Anthropic spoke at the Nonprofit Leadership Conference about AI's accelerating pace. Tiff Macklem of the Bank of Canada says AI will displace workers in certain professions. Kristalina Georgieva of the IMF notes that service jobs pay little and that new job creation is hard to predict. Ayesha Kahnna of Addo AI highlights Singapore's subsidized skills training programs.
Small businesses lead the AI economy
A Fast Company survey of Goldman Sachs 10,000 Small Businesses alumni shows 86% now use AI. However only 15% have fully integrated AI into their operations. Pompey compared early AI adoption to blacksmiths who adapted to automobiles. Small businesses are increasingly using AI for customer queries and financial analysis.
Google launches free offline AI training for Nigerians
Google launched a free learning platform called Skilling Africa with Google for Nigerians. It offers over 15 digital and AI courses, some of which can be completed offline. The platform aims to train three million Africans by 2030. Courses cover topics like AI, cloud computing, and digital marketing. It also includes an AI study assistant called SCOUT.
How to tell if a voice recording is AI-generated
Advances in AI voice cloning make it easy to copy a voice in seconds. Researchers say our ears are often bad at telling real voices from cloned ones. Cloned voices used to sound too polished, but now they include fake breaths and pauses. Scientists use audio forensics to study sound wave features like jitter and shimmer to detect fakes. However, no single tool can catch all AI-generated voices.
Sources
- ChatGPT Is Giving Voters Ludicrous Misinformation Right as the Midterms Hit
- What happens when voters turn to AI chatbots for election information?
- Some voters are using AI chatbots for Election Day guidance
- As the controversy around AI-created video game copies continues, Anthropic seemingly restricts Claude's cloning abilities
- Claude’s new rules against game cloning won’t stop the AI copying boom
- China’s robot makers have mastered the moves. Now comes the race for brains.
- Helpful until we’re helpless? Researchers warn AI is gradually eroding human control
- Why mandating AI use in the workplace can backfire
- Tackling AI's Labor Paradox
- The AI economy is a small-business economy
- Google launches free offline AI, digital skills training for Nigerians
- How to tell if a voice recording is AI-generated
Comments
Please log in to post a comment.