Google's AI Mode Avoids Naming Biden's 2020 Win

Key Takeaways

• <ul><li>Google's AI Mode avoids stating Joe Biden won the 2020 election, directing users to FEC documents instead</li><li>Google has not commented on whether the Trump administration influenced the chatbot's response</li><li>Anthropic filed for IPO under ticker ANT with a potential $2 trillion valuation</li><li>Anthropic currently runs at $67 billion annually with an 11% profit margin</li><li>Anthropic faces a lawsuit over 274 Elvis Presley compositions, accused of using lyrics without permission</li><li>Elon Musk warns of neocloud sector collapse despite rebounds by CoreWeave, Nebius, and Iren</li><li>AI-generated celebrity digital twins increase customer trust and forgiveness of mistakes, per Iowa State University research</li><li>AI resume filters in Indianapolis are rejecting qualified applicants, prompting recommendations to tailor resumes with keywords</li><li>LogicalDOC 9.3.1 adds AI summarization using Maximal Marginal Relevance and NLP techniques</li><li>Process-aware knowledge graphs reduced invalid API calls from 46% to 3% in LLM agent testing</li></ul>

Google AI Mode won't say who won the 2020 election

Google's AI Mode avoids answering who won the 2020 presidential election. Instead of stating Joe Biden won, it directs users to check Federal Election Commission documents. Google said it is working to fix the issue. A Google spokesman did not comment on whether the Trump administration influenced the chatbot's response.

Google chatbot deflects question about 2020 election winner

Google's AI chatbot refused to directly say Joe Biden won the 2020 presidential election. It told users to review official FEC documents showing Biden earned 306 Electoral College votes to Trump's 232. The chatbot also added a warning that AI can make mistakes. A Google spokesman did not answer questions about possible involvement from the Trump administration.

Google AI gives evasive answer about 2020 election winner

Google's AI Mode gives evasive answers when asked who won the 2020 presidential election. Instead of acknowledging Joe Biden's win, it tells users to consult outside sources. The AI's response contrasts with how some US politicians discuss the election results. The article notes that direct questioning produces a cagey reaction from the AI tool.

AI-generated celebrity faces affect customer trust

A study co-authored by Iowa State University researcher Lingyao (Ivy) Yuan found that AI-generated celebrity digital twins shape customer perceptions. People are more likely to trust a service and forgive its mistakes when a virtual representative resembles a famous person they like. The research shows a halo effect from positive feelings about the celebrity carrying over to the service.

Study reveals impact of celebrity AI digital twins

A study by Iowa State University's Lingyao (Ivy) Yuan and researchers from Indiana University and the University of Sydney examined AI-powered digital customer service representatives. The team used Hugh Jackman's face on a digital representative to test customer reactions. Participants who interacted with the celebrity version were more trusting and more forgiving of mistakes. Yuan said the findings point to a future where digital humans become routine in everyday business.

Retrieving Facts Is Not Enough When Obsolete Data Stays in Agent Memory

Researchers tested AI memory agents using Qwen2.5-7B-Instruct on the FactConsolidation task. Accuracy dropped when more evidence was retrieved, even though the newest facts were present. Obsolete competing statements caused harmful errors more than unrelated evidence of the same size. The study shows that finding the right fact is not enough if outdated versions remain in memory.

LogicalDOC 9.3.1 Adds AI Summarization and Infrastructure Updates

LogicalDOC released version 9.3.1 with a new Summarization feature powered by AI and NLP techniques. The system uses methods like Maximal Marginal Relevance to generate concise document summaries. The update also includes Java 25 LTS and improvements to the ORM layer for better database access. The release aims to make document content easier to understand and use for businesses.

AI Resume Filters May Be Rejecting Qualified Applicants, Here Is How to Fix That

Job seekers in Indianapolis report being rejected by AI resume screening systems despite being qualified. WorkOne recommends tailoring resumes to each job posting by matching keywords from the description. The AI tool Hardly scores resumes and applicants should aim for 80% to 90%. WorkOne Indy hosts free resume workshops each Thursday, and the Indianapolis Public Library also offers related sessions.

Anthropic IPO Valuation Debate Breaks Down to Simple Math

Anthropic filed to go public under the ticker ANT with a potential $2 trillion valuation. The valuation depends on 2028 revenue, profit margin, and profit multiple predictions. Anthropic currently runs at $67 billion a year with an 11% profit margin. The calculator compares Anthropic against companies like Meta, Netflix, and Nvidia to estimate whether the $2 trillion top range is reasonable.

Elvis Rights Holders Sue Anthropic Over AI Training and Lyrics Output

ABG EPE IP LLC and other rights holders filed a lawsuit against Anthropic over 274 Elvis Presley compositions. The complaint accuses Anthropic of collecting lyrics from sources like Common Crawl and The Pile without permission. It also alleges Anthropic used BitTorrent to download pirated book libraries and removed copyright information during data cleaning. The plaintiffs seek a permanent injunction and destruction of unauthorized copies.

Elon Musk Warns of Neocloud Collapse as Rivals Rebound

Elon Musk warned that the neocloud sector faces a massive wave of collapses due to debt-fueled expansions. Despite this prediction, companies like CoreWeave, Nebius and Iren are showing resilience with recent rebounds. Data center companies including Vertiv are also part of the sector's ongoing developments.

Alfieholdingslimited Launches AI Financial Analytics Platform

Alfieholdingslimited launched an AI-powered market intelligence and financial analytics platform in Brussels on October 28, 2026. The company uses AI to analyze market data, identify patterns and support research workflows. It emphasizes data security through encryption, access controls and audit logging. The platform is not a brokerage or investment service and stresses transparency about AI limitations.

Process-Aware Knowledge Graphs Improve LLM Agent Reliability

Researchers developed a method using Open Data Protocol metadata and Business Process Model and Notation in a typed property graph to improve LLM agent reliability. The approach improved Recall at 3 by 5.7 points for queries with low lexical overlap across 42 enterprise services. An LLM agent completed 38.7 percent of multi-step tasks compared to 10.3 percent with a baseline method. Full typed context reduced invalid API calls from 46 percent to just 3 percent.

Sources

• Google is aware its AI sometimes won't say who won the 2020 election
• Google’s AI Chatbot Refuses to Answer Question of Who Won 2020 Presidential Election
• Google AI gets cagey when asked who won the 2020 elections
• When artificial intelligence wears a famous face
• When Artificial Intelligence Wears Famous Face
• When Retrieval Is Not Enough: Obsolete Facts in Agent Memory
• LogicalDOC 9.3.1: Artificial Intelligence Simplifies Access to Business Document Content
• AI filters may be screening out your resume. Here’s how to help.
• Etzioni on AI: Is Anthropic worth $2 trillion? Do the napkin math with this handy calculator
• Rights Holders of Elvis Compositions File Suit Against Anthropic Over AI Training and Lyrics Output
• Cloud Stock Rebound Continues, Even As Elon Musk Warns Of 'Guppy-Level Death'
• Alfieholdingslimited Launches AI-Powered Market Intelligence and Financial Analytics Platform
• Grounding LLM Agents in Business Processes: Process-Aware Knowledge Graphs for Reliable Multi-Step Service Invocation