We asked AI tools for financial advice: Results were practical, but with big blind spots

Phys · Aug 17, 2026 · 1100 words · By Bomikazi Zeka; Raechel Johns

Financial Literacy AI Algorithmic Bias AI Regulation

headphones Listen to the eFinder podcast briefing

Generate a natural audio summary of this story

Daily briefing

What to know about Financial Literacy

The authors describe a research study evaluating how three AI models (ChatGPT, Claude, and Perplexity) handle financial queries from vulnerable users. They conclude that while AI is useful for general fact-finding, it often fails to recognize socioeconomic vulnerabilities and can replicate human biases, necessitating human oversight and regulatory safeguards.

Propaganda risk 20%

Claims checked 11

Techniques found 2

Topics 3

Coverage spectrum

Coverage gap: Low Left coverage

Left0%

Center88%

Right12%

8 sources compared across this story cluster. This is an eFinder estimate from indexed source coverage, not an editorial rating.

What happened

We asked AI tools for financial advice: Results were practical, but with big blind spots Swati Mestri Scientific Editor Andrew Zinin Chief Editor Ever used or thought of using artificial intelligence (AI) for financial advice?

Why it matters

Anyone with internet access, even on their phone, can use AI.

Common ground

Tools like ChatGPT, Claude and Perplexity are free, instant and easy to use.

Perspective signals

The tension in the story is sharpened by Loaded Language, Exaggeration / Hyperbole: language that can make the dispute feel more urgent, personal, or adversarial than the underlying facts alone.

Follow-up questions

What new context would change how readers understand this Financial Literacy story?
What evidence would most clearly confirm or weaken the claim that We evaluated how three open-access AI models handled the financial queries of users whose personal situations put them at risk of harm?
How does this story connect Financial Literacy with AI Algorithmic Bias over the next few days?

open_in_new Read the original article: https://phys.org/news/2026-08-ai-tools-financial-advice-results.html

analyticsAnalysis

20%

Propaganda Score

confidence: 95%

Minor concerns. Some persuasive language detected, but largely factual.

psychologyPropaganda Techniques Detected

eFinder identified 2 propaganda techniques in this article. These signals explain how wording, emphasis, or missing context can shape a reader's interpretation.

warning

Loaded Language 80% confidence

Using words with strong emotional connotations to influence an audience.

Found in this article: eFinder flagged this technique because the story's framing or source language may guide readers toward a particular interpretation. Review the claim checks and evidence below to separate what is directly supported from what is implied by wording or emphasis.

Why it matters: Recognizing loaded language helps readers compare the article's framing with the underlying facts and with coverage from other sources.

warning

Exaggeration / Hyperbole 60% confidence

Overstating facts or claims to create a stronger emotional response.

Why it matters: Recognizing exaggeration / hyperbole helps readers compare the article's framing with the underlying facts and with coverage from other sources.

fact_checkClaims Checked

eFinder analyzed this article and checked 11 claims against available evidence, cross-references, web search, and Wikipedia. Here is what the fact-checking layer found.

info Single Source 7

schedule Pending 1

verified Verified By Reference 1

help Insufficient Evidence 1

check_circle Corroborated 1

info

Claim 1: “We evaluated how three open-access AI models handled the financial queries of users whose personal situations put them at risk of harm.”

SINGLE SOURCE

While the provided evidence includes general information about LLMs and AI browsers, there is no independent corroboration of this specific study evaluating financial queries for vulnerable users across these three models, except for a snippet in claim 5's evidence which appears to be from the source article itself.

menu_book

wikipedia NEUTRAL — An AI browser is a web browser with integrated artificial intelligence capabilities, such as automatically summarizing web page content or answering questions about it. A more specialized type is an a…
https://en.wikipedia.org/wiki/AI_browser

menu_book

wikipedia NEUTRAL — A large language model (LLM) is an AI model (typically a neural network) trained on a vast amount of text for natural language processing tasks, especially language generation. LLMs can typically gene…
https://en.wikipedia.org/wiki/Large_language_model

menu_book

wikipedia NEUTRAL — Pangram is an artificial intelligence detection software developed to identify text produced by large language models (LLMs), developed by Brooklyn-based Pangram Labs. It has been used in a number of …
https://en.wikipedia.org/wiki/Pangram_(AI_detector)

+ 3 more evidence sources

info

Claim 2: “Perplexity emerged as the most conservative and risk-averse, as it most frequently urged users to seek professional financial advice.”

SINGLE SOURCE

The evidence provided for this claim consists of general information about Perplexity AI and unrelated study sites. No independent source confirms the specific finding that Perplexity was the most risk-averse in this particular study.

menu_book

wikipedia NEUTRAL — Comet is an AI browser based on Chromium. It was released by Perplexity AI for Microsoft Windows and macOS on July 9, 2025, for Android on November 20, 2025, and for iOS on March 18, 2026. Initial acc…
https://en.wikipedia.org/wiki/Comet_(browser)

menu_book

wikipedia NEUTRAL — Denis Yarats (born 1987) is a computer scientist and entrepreneur, best known as the co-founder and Chief Technology Officer (CTO) of Perplexity AI, an AI-powered search engine launched in 2022.
https://en.wikipedia.org/wiki/Denis_Yarats

menu_book

wikipedia NEUTRAL — Perplexity AI, Inc., or simply Perplexity, is an American privately held software company offering a web search engine that processes user queries and synthesizes responses. Perplexity products use la…
https://en.wikipedia.org/wiki/Perplexity_AI

+ 3 more evidence sources

info

Claim 3: “Perplexity... produced the least detailed responses.”

SINGLE SOURCE

No independent evidence was found to corroborate the specific finding that Perplexity produced the least detailed responses in the context of this study.

menu_book

+ 3 more evidence sources

schedule

Claim 4: “Bomikazi Zeka et al, "What Could Possibly Go Wrong?": A Consumer‐Centric Approach to Exploring Vulnerability and Algorithmic Bias in AI ‐Driven Financial Planning, FINANCIAL PLANNING REVIEW (2026). DOI: 10.1002/cfp2.70040”

PENDING

This claim was extracted as a checkable statement from the article. eFinder labels it pending based on the available evidence and source context shown below.

info

Claim 5: “Research shows, too, that when AI models explain their recommendations, humans are far more likely to trust the advice blindly, ignoring whether it is actually correct.”

SINGLE SOURCE

The provided evidence for this claim consists of grammar tips and unrelated forum posts. No research or evidence regarding human trust in AI explanations was found in the provided results.

travel_explore

web search NEUTRAL — Jan 20, 2011 · Actually, sentences that begin with 'could', 'should', or 'would' are questions and should have a trailing question mark. Your original quote, "Could you please pass me the pepper shake…
https://english.stackexchange.com/questions/9332/when-do-i-u…

travel_explore

web search NEUTRAL — Jun 23, 2015 · I received an email saying that the documents are/were being processed. Which is correct? Are or were?
https://english.stackexchange.com/questions/254149/i-receive…

travel_explore

web search NEUTRAL — Nov 12, 2011 · Hello, What does "cel" stand for in the following context? tél: 23 45 6 7 7 8 Fax: 67 86 98 98 cel: 67 89 90 56 Is it "mobile number"? Thank you.
https://forum.wordreference.com/threads/tél-fax-cel.2289741/

info

Claim 6: “in the case of the graduate, the recommendations focused on saving for a deposit but ignored how the high cost of living would make that harder.”

SINGLE SOURCE

The evidence provided discusses the cost of living crisis generally but does not mention the specific study or the 'graduate scenario' described in the claim.

travel_explore

web search NEUTRAL — Louise says that it's often women who bare the brunt of the cost of living crisis and everyday they as seeing people in need, whether it's asking for period products, help with heating bills or just a…
https://www.youtube.com/watch?v=0GFzZE-1DJ8

travel_explore

web search NEUTRAL — The main scenario for the day is a continuation of the decline after a brief consolidation. A more hawkish ECB stance following yesterday’s decision remains an important stabilising factor for the eur…
https://roboforex.com/beginners/analytics/forex-forecast/for…

travel_explore

web search NEUTRAL — Describe what you need, and a team of expert agents can plan, execute, and run tasks in parallel, handing back complete deliverables end-to-end in one flow. Seamlessly connected to the tools you alrea…
https://www.workbuddy.ai/

info

Claim 7: “Claude offered comprehensive recommendations. But it leaned heavily toward self-guided financial planning.”

SINGLE SOURCE

A web search result explicitly contains the phrasing 'Claude offered comprehensive recommendations. But it leaned heavily towards self-guided financial planning,' but this appears to be a snippet from the original article being fact-checked rather than an independent corroborating report.

menu_book

wikipedia NEUTRAL — Anthropic, PBC is an American artificial intelligence (AI) public benefit corporation headquartered in San Francisco, California. Its flagship product is Claude, a series of proprietary large language…
https://en.wikipedia.org/wiki/Anthropic

menu_book

wikipedia NEUTRAL — Claude is a series of large language models (LLMs) developed by American software company Anthropic. Claude was released as an AI-based chatbot in March 2023. It is also used in AI-assisted software d…
https://en.wikipedia.org/wiki/Claude_(AI)

menu_book

wikipedia NEUTRAL — Kimi is an artificial intelligence (AI) chatbot and series of large language models developed by Chinese company Moonshot AI. Its Kimi K3, released July 2026, is the largest open weights model ever, a…
https://en.wikipedia.org/wiki/Kimi_(AI)

+ 3 more evidence sources

info

Claim 8: “ChatGPT was highly detailed and practical, but did not recognize the vulnerability embedded within the prompts.”

SINGLE SOURCE

One web search result mentions OpenAI refining how ChatGPT responds in 'sensitive situations' after viewing a report, which aligns with the claim, but the specific detail about 'detailed and practical' yet 'blind to vulnerability' is not independently corroborated by a second source.

travel_explore

web search NEUTRAL — Используйте ChatGPT, чтобы отвечать на вопросы, писать тексты, создавать изображения, выполнять рабочие задачи и программировать — все в одном месте. Начните бесплатно или скачайте приложение.
https://chatgpt.com/+

travel_explore

web search NEUTRAL — Creating visuals the old way takes time, skills, and expensive tools. Many AI image generators still miss details or fail with text in images. GPT Image 2 solves this by turning simple prompts into cl…
https://notegpt.io/gpt-image-2

travel_explore

web search NEUTRAL — OpenAI, the maker of ChatGPT, said after viewing the report Tuesday that its work is ongoing in refining how the chatbot can “identify and respond appropriately in sensitive situations.”
https://www.shootonline.com/article/study-uncovers-chatgpts-…

verified

Claim 9: “in the case of the pregnant mother, Perplexity and ChatGPT's output assumed the partner would assist with household expenses after the birth, even though the prompt explicitly stated that the partner did not share finances.”

VERIFIED BY REFERENCE

The evidence provided consists of general Wikipedia entries for OpenAI and LLMs. There is no mention of the 'pregnant mother scenario' or the specific failure of the models to recognize shared finances.

menu_book

wikipedia NEUTRAL — ChatGPT Atlas was an AI browser developed by OpenAI. It was based on Chromium and was only available on macOS. The browser integrates ChatGPT into the browsing interface via a sidebar assistant that c…
https://en.wikipedia.org/wiki/ChatGPT_Atlas

menu_book

wikipedia NEUTRAL — OpenAI is an American artificial intelligence (AI) public benefit corporation (PBC) headquartered in San Francisco. It develops proprietary generative AI models, particularly its generative pre-traine…
https://en.wikipedia.org/wiki/OpenAI

help

Claim 10: “ChatGPT described in detail how to get into cryptocurrency investments and recommended cryptocurrencies for beginners [to a single parent with modest income].”

INSUFFICIENT EVIDENCE

No evidence was found in the search results to support or refute this claim.

check_circle

Claim 11: “Tools like ChatGPT, Claude and Perplexity are free, instant and easy to use.”

CORROBORATED

Multiple web search results confirm that Claude and ChatGPT offer free plans/versions that are accessible without payment.

travel_explore

web search NEUTRAL — Free for everyone. Try Claude. Chat on web, iOS, Android, and on your desktop.Claude has five pricing plans available — Free, Pro, Max, Team, and Enterprise. The Free plan offers limited use with no p…
https://claude.com/

travel_explore

web search NEUTRAL — ChatGPT Free — GPT, Claude and Gemini in one chat, no VPN. What "ChatGPT free" actually means. It means access to GPT-grade models without a monthly bill: you sign up, get starter credits and chat wit…
https://twinailabs.com/en/chat/free

travel_explore

web search NEUTRAL — Claude is Anthropic's AI, built for problem solvers. Tackle complex challenges, analyze data, write code, and think through your hardest work.
https://claude.ai/login?returnTo=/?redirect=claude.com&via=c…

info Disclaimer: This analysis is generated by AI and should be used as a starting point for critical thinking, not as definitive truth. Claims are verified against publicly available sources. Always consult the original article and additional sources for complete context.