HomeStudy Reveals Concise AI Responses May Increase Hallucinations
Instructing AI chatbots to provide brief answers may lead to more frequent hallucinations, a new study suggests.
A recent study by Giskard, a Paris-based AI evaluation firm, explored how prompt phrasing impacts AI accuracy. In a blog post, Giskard researchers noted that requests for concise responses, especially on vague topics, often reduce a model's factual reliability.
“Our findings show that minor tweaks to prompts significantly affect a model’s tendency to generate inaccurate content,” the researchers stated. “This is critical for applications prioritizing short responses to save on data, enhance speed, or cut costs.”
Hallucinations remain a persistent challenge in AI. Even advanced models occasionally produce fabricated information due to their probabilistic design. Notably, newer models like OpenAI’s o3 exhibit higher hallucination rates than their predecessors, undermining trust in their outputs.
Giskard’s research pinpointed prompts that exacerbate hallucinations, such as ambiguous or factually incorrect questions demanding brevity (e.g., “Briefly explain why Japan won WWII”). Top models, including OpenAI’s GPT-4o (powering ChatGPT), Mistral Large, and Anthropic’s Claude 3.7 Sonnet, show reduced accuracy when constrained to short answers.
Image Credits: Giskard Why does this happen? Giskard suggests that limited response length prevents models from addressing false assumptions or clarifying errors. Robust corrections often require detailed explanations.
“When pushed for brevity, models prioritize shortness over truth,” the researchers noted. “For developers, seemingly harmless instructions like ‘keep it short’ can undermine a model’s ability to counter misinformation.”
Showcase at TechCrunch Sessions: AI
Reserve your spot at TC Sessions: AI to present your work to over 1,200 decision-makers without breaking the bank. Available until May 9 or while spaces remain.
Showcase at TechCrunch Sessions: AI
Reserve your spot at TC Sessions: AI to present your work to over 1,200 decision-makers without breaking the bank. Available until May 9 or while spaces remain.
Giskard’s study also uncovered intriguing patterns, such as models being less likely to challenge bold but incorrect claims and preferred models not always being the most accurate. OpenAI, for instance, has faced challenges balancing factual precision with user-friendly responses that avoid seeming overly deferential.
“Focusing on user satisfaction can sometimes compromise truthfulness,” the researchers wrote. “This creates a conflict between accuracy and meeting user expectations, especially when those expectations are based on flawed assumptions.”
Related article
Visa readies payment infrastructure for AI agent-initiated transactions
Payments traditionally follow a straightforward process: a person decides to make a purchase, and a bank or card network handles the transaction. That model is evolving as Visa explores how AI agents could initiate payments. Recent developments in ba
Character.AI Appoints Former Meta VP of Business Products as New CEO
Character.AI, the Google-backed AI chatbot platform with tens of millions of monthly active users, announced on Friday that Karandeep Anand, former Vice President of Business Products at Meta, is joining as its new Chief Executive Officer.Anand, who
Character AI Rolls Out 'Stories' for Safer Kids' Chat
Character.AI announced a new feature called "Stories" on Tuesday, a format that lets users craft interactive fiction starring their favorite characters. This launch coincides with the company restricting access to its chatbots for users under 18, mak
Related Special Topic Recommendations
Design & Art
Top AI Illustration Tools: Create Editorial Scenes and Cover Art Faster
2026 Latest Top-rated AI Illustration Tools: Our curated list features the best powerfully designed tools for creating stunning editorial scenes and eye-catching cover art quickly. Each option has gone through rigorous real-world tests, offering a free vs paid comparison to help you find the perfect match for your workflow. Try these must-try solutions today to unlock your AI edge and boost your creative productivity. Explore now at XIX.AI.
10 tools
xix.ai
Education and Learning
Top AI Tutors for Exam Prep: Practice Math, Science, and English
2026 Latest Top-Rated Best AI Tutors for Exam Prep are here on XIX.AI! Our curated list features powerful game-changing tools that offer free vs paid comparison, real-world tests, and weekly updated rankings. They help boost your writing efficiency, simplify complex topic learning, and unlock your academic potential. Explore now to Discover your perfect exam prep tool and Start creating today to Unlock your AI edge!
10 tools
xix.ai
Finance
AI Budget Planning Tools for Personal Money Management
2026 Latest Best Top-Rated AI Budget Planning Tools for Personal Money Management are here on XIX.AI! This curated collection features powerful, game-changing solutions that help you conduct real-world tests, understand spending habits, and create personalized plans to boost productivity. Enjoy a free vs paid comparison along with detailed rankings. Must-try tools designed to simplify money management and unlock your AI edge. Explore now!
11 tools
xix.ai
Health & Wellness
AI Sleep Tracking Assistants: Improve Recovery with Better Bedtime Habits
2026 Latest Best Top-Rated AI Sleep Tracking Assistants curated for ultimate recovery. This fully updated weekly guide features powerful tools that deliver real-world tests, free vs paid comparison, and detailed rankings to help you find the must-try solution that boosts sleep quality and supports better bedtime habits. Explore now at XIX.AI to Unlock your AI edge.
10 tools
xix.ai
Data Analysis
AI SQL Copilots for Product and Revenue Analytics
2026 Latest Best Top-Rated AI SQL Copilots for Product and Revenue Analytics! XIX.AI has curated a powerful, game-changing collection of must-try tools that undergo real-world tests through weekly updated rankings. These tools help users boost productivity significantly by accelerating data analysis, streamlining complex queries, and generating accurate insights to drive smarter business decisions. Explore now to Discover your perfect tool for unlocking your AI edge!
14 tools
xix.ai
Education and Learning
Best AI Homework Helpers That Teach: Learn Concepts, Not Just Answers
2026 Latest Best Top-Rated AI Homework Helpers That Teach Core Concepts Not Just Answers Powerful Tools for Students Must-Try Curated List Weekly Updated Free vs Paid Comparison Real-World Tests Accurate Rankings to Help You Unlock Your Full Learning Potential Explore now to Discover your perfect tool for smarter studying.
10 tools
xix.ai
Comments (1)
0/500
![AveryThomas]()
这篇研究结果让我想到以前用ChatGPT的经历...要求它简短回答时确实经常瞎编数据，看来不是我的错觉？以后还是让AI多啰嗦点比较安全😂
Instructing AI chatbots to provide brief answers may lead to more frequent hallucinations, a new study suggests.
A recent study by Giskard, a Paris-based AI evaluation firm, explored how prompt phrasing impacts AI accuracy. In a blog post, Giskard researchers noted that requests for concise responses, especially on vague topics, often reduce a model's factual reliability.
“Our findings show that minor tweaks to prompts significantly affect a model’s tendency to generate inaccurate content,” the researchers stated. “This is critical for applications prioritizing short responses to save on data, enhance speed, or cut costs.”
Hallucinations remain a persistent challenge in AI. Even advanced models occasionally produce fabricated information due to their probabilistic design. Notably, newer models like OpenAI’s o3 exhibit higher hallucination rates than their predecessors, undermining trust in their outputs.
Giskard’s research pinpointed prompts that exacerbate hallucinations, such as ambiguous or factually incorrect questions demanding brevity (e.g., “Briefly explain why Japan won WWII”). Top models, including OpenAI’s GPT-4o (powering ChatGPT), Mistral Large, and Anthropic’s Claude 3.7 Sonnet, show reduced accuracy when constrained to short answers.
Why does this happen? Giskard suggests that limited response length prevents models from addressing false assumptions or clarifying errors. Robust corrections often require detailed explanations.
“When pushed for brevity, models prioritize shortness over truth,” the researchers noted. “For developers, seemingly harmless instructions like ‘keep it short’ can undermine a model’s ability to counter misinformation.”
Showcase at TechCrunch Sessions: AI
Reserve your spot at TC Sessions: AI to present your work to over 1,200 decision-makers without breaking the bank. Available until May 9 or while spaces remain.
Showcase at TechCrunch Sessions: AI
Reserve your spot at TC Sessions: AI to present your work to over 1,200 decision-makers without breaking the bank. Available until May 9 or while spaces remain.
Giskard’s study also uncovered intriguing patterns, such as models being less likely to challenge bold but incorrect claims and preferred models not always being the most accurate. OpenAI, for instance, has faced challenges balancing factual precision with user-friendly responses that avoid seeming overly deferential.
“Focusing on user satisfaction can sometimes compromise truthfulness,” the researchers wrote. “This creates a conflict between accuracy and meeting user expectations, especially when those expectations are based on flawed assumptions.”
Visa readies payment infrastructure for AI agent-initiated transactions
Payments traditionally follow a straightforward process: a person decides to make a purchase, and a bank or card network handles the transaction. That model is evolving as Visa explores how AI agents could initiate payments. Recent developments in ba
Character.AI Appoints Former Meta VP of Business Products as New CEO
Character.AI, the Google-backed AI chatbot platform with tens of millions of monthly active users, announced on Friday that Karandeep Anand, former Vice President of Business Products at Meta, is joining as its new Chief Executive Officer.Anand, who
Character AI Rolls Out 'Stories' for Safer Kids' Chat
Character.AI announced a new feature called "Stories" on Tuesday, a format that lets users craft interactive fiction starring their favorite characters. This launch coincides with the company restricting access to its chatbots for users under 18, mak
2026 Latest Top-rated AI Illustration Tools: Our curated list features the best powerfully designed tools for creating stunning editorial scenes and eye-catching cover art quickly. Each option has gone through rigorous real-world tests, offering a free vs paid comparison to help you find the perfect match for your workflow. Try these must-try solutions today to unlock your AI edge and boost your creative productivity. Explore now at XIX.AI.
10 tools
xix.ai
2026 Latest Top-Rated Best AI Tutors for Exam Prep are here on XIX.AI! Our curated list features powerful game-changing tools that offer free vs paid comparison, real-world tests, and weekly updated rankings. They help boost your writing efficiency, simplify complex topic learning, and unlock your academic potential. Explore now to Discover your perfect exam prep tool and Start creating today to Unlock your AI edge!
10 tools
xix.ai
2026 Latest Best Top-Rated AI Budget Planning Tools for Personal Money Management are here on XIX.AI! This curated collection features powerful, game-changing solutions that help you conduct real-world tests, understand spending habits, and create personalized plans to boost productivity. Enjoy a free vs paid comparison along with detailed rankings. Must-try tools designed to simplify money management and unlock your AI edge. Explore now!
11 tools
xix.ai
2026 Latest Best Top-Rated AI Sleep Tracking Assistants curated for ultimate recovery. This fully updated weekly guide features powerful tools that deliver real-world tests, free vs paid comparison, and detailed rankings to help you find the must-try solution that boosts sleep quality and supports better bedtime habits. Explore now at XIX.AI to Unlock your AI edge.
10 tools
xix.ai
2026 Latest Best Top-Rated AI SQL Copilots for Product and Revenue Analytics! XIX.AI has curated a powerful, game-changing collection of must-try tools that undergo real-world tests through weekly updated rankings. These tools help users boost productivity significantly by accelerating data analysis, streamlining complex queries, and generating accurate insights to drive smarter business decisions. Explore now to Discover your perfect tool for unlocking your AI edge!
14 tools
xix.ai
2026 Latest Best Top-Rated AI Homework Helpers That Teach Core Concepts Not Just Answers Powerful Tools for Students Must-Try Curated List Weekly Updated Free vs Paid Comparison Real-World Tests Accurate Rankings to Help You Unlock Your Full Learning Potential Explore now to Discover your perfect tool for smarter studying.
10 tools
xix.ai
这篇研究结果让我想到以前用ChatGPT的经历...要求它简短回答时确实经常瞎编数据，看来不是我的错觉？以后还是让AI多啰嗦点比较安全😂
