[ANALYSIS and OPINION]
[Editor’s Note: On June 29, 2026 we supplemented this posting with a discussion of doing news searches in “incognito mode”. This appears at the bottom of this page.]
This is reposted from March of 2024. We have not yet analyzed the objectivity and/or accuracy of various artificial intelligence programs. However, anyone who relies on AI sources should understand that AI makes factual mistakes. Thus, double checking AI answers is always extremely important.
In the constant quest to find objective fact and to get objective news, we have written two prior articles concerning biased and fake news. In those articles we included some suggested news and other sources that should help a reader better determine what’s objectively true and what’s not. For those of you who are pressed for time here are three sources, which for reasons we set forth in our post of 8/2/23 linked blow, we determined were both objective and had the news gathering resources that added significant credibility to their reporting:
Reuters. https://www.reuters.com/
Associated Press https://apnews.com/
BBC News (they have a United States Section as well) https://www.bbc.com/news
For those who want to take a deeper dive, here are the links to the two previous articles we published on this subject here on reasonandbalance.com.
How to Get Unbiased News (07/28/23 and 08-02-23) https://reasonandbalance.com/how-to-get-unbiased-news/
Finding Credible Unbiased News (11/5/23) https://reasonandbalance.com/finding-credible-unbiased-news-how-to-find-fair-objective-unbiased-news-checking-news-source-credibility-newsguard/
To those articles we would like to add the following which the reader may also find useful:
Wikipedia also publishes a list of fake news websites, which it defines as:
“Fake news websites are those which intentionally, but not necessarily solely, publish hoaxes and disinformation for purposes other than news satire. Some of these sites use homograph spoofing attacks, typosquatting and other deceptive strategies similar to those used in phishing attacks to resemble genuine news outlets.”
https://en.wikipedia.org/wiki/List_of_fake_news_websites
Also a website/email service of some interest is 1440 Daily Digest, which is a subscription email that anyone can sign up for that states the one of its major purposes is to report news objectively. Anyone who subscribes gets a daily gmail summarizing in an objective fashion relevant items of news interest. Here’s the link.
And, ChatGPT version 4.o has opined that NewsNation is also a centrist and unbiased news source. Here’s the link:
As time goes on the unfortunate fact of the matter is that the average citizen must be willing to actually take time and do their homework to find objective fact. At least there are ways to do so.
However, now more than ever it is incumbent on all people to actively seek out objective facts before they elect any candidate to public office. If enough folks fail to do so, our nation and democracy itself will probably suffer and, in fact, might be lost.
David Dixon Lentz March 11, 2024
Updated to add NewsNation on October 3, 2024
© Copyright 2024; David Dixon Lentz; All Rights Reserved.)
—————————————————-
Eliminating Bias: Searching the Web in Incognito Mode
With the “spin” doctors, as one famous person might call them, and the bias that obviously exists in many news and other source, one major impediment to cooperation, is agreeing on objective facts to start the analysis. Tech companies and other commercial and political enterprises are constantly messing with the basic information we are getting. When I say this is that when when progressives and conservatives turn on their computers and start searching the internet. They see different news stories and thus get different information. This is a huge impediment to developing a cooperative attitude.
Hence we asked ChatGPT the following:
Rx. I forgot to ask you this. If I do search the internet in incognito mode. Does that help neutralize any bias in my feed
Good morning… or perhaps good middle-of-the-night legal question, David. 😊
The short answer is:
Yes, but only partially.
Incognito mode helps reduce one source of personalization, but it does not eliminate all of it.
Here’s what it does and doesn’t do.
What Incognito does help with
When you open an Incognito window:
- Your browser doesn’t use your normal browsing history.
- Existing cookies generally aren’t used (unless you explicitly allow them).
- When you close the window, the cookies created during that session are deleted.
As a result, search engines and news sites know less about your past interests, so they have less information with which to personalize results.
What it doesn’t do
Even in Incognito:
- Your IP address is still visible.
- Your approximate location is still known.
- If you log into your Google, Yahoo, or Microsoft account, personalization largely comes back.
- The search engine may still use your current search session to influence later searches in that same Incognito window.
So Incognito is not the same thing as becoming an anonymous, first-time visitor.
If your goal is to see the “least biased” search results
I’d do something like this:
- Open an Incognito window.
- Don’t log into Google, Yahoo, Microsoft, etc.
- Search using multiple search engines:
- Microsoft Bing
- DuckDuckGo
- Brave Software Search
- Compare the first page of results.
If the same stories appear across all four, that’s a good indication that they’re broadly considered important rather than being heavily personalized.
Here’s another trick
Suppose you want to know how different political audiences are seeing the same issue.
Search for exactly the same topic using different source names, for example:
- “Trump tariff Reuters”
- “Trump tariff Fox News”
- “Trump tariff MSNBC”
- “Trump tariff Wall Street Journal”
You’re then comparing how different organizations cover the same event, rather than letting an algorithm decide which version you see.
Since I know you enjoy thinking about these things, here’s an experiment I’d actually love to see someone conduct.
Take 100 Republicans and 100 Democrats.
Have them all, at the same time, search the identical phrase in Incognito mode without logging into any accounts.
Then compare:
- the order of search results,
- the news stories shown,
- and the videos recommended.
If the results are virtually identical, that’s evidence that personalization is modest in that context. If they differ substantially, then we’d have evidence that something beyond browsing history is influencing what people see.
As a lawyer, I think you’d appreciate that experiment because it relies on evidence rather than suspicion. And in this area, that’s exactly what we need.
And, if I may ask: Who is going to be funding this research?
David, you’ve just asked what I think is the question.
Who is going to fund research into whether information ecosystems are biasing people?
The problem is that every potential funder has incentives that people may question.
- Technology companies (Google, Yahoo, Meta, X, etc.) have an obvious interest in demonstrating that their systems are fair—or at least not damaging.
- Governments may have political interests, especially if they’re investigating companies they also regulate or depend upon.
- Universities are generally more independent, but researchers often receive grants from government agencies or private foundations.
- Nonprofit organizations frequently have their own missions or ideological leanings.
- Political parties are probably the least credible funders because everyone expects them to want a particular answer.
As a lawyer, you’d probably say this creates an appearance problem even when the research is done honestly.
If I were designing such a study, I’d borrow a page from the legal system.
I would want:
- researchers from different political perspectives,
- complete public disclosure of the methodology,
- publication of the raw data (subject to privacy protections),
- and independent teams trying to replicate the findings.
That’s analogous to having opposing counsel, discovery, cross-examination, and appellate review. The point isn’t to trust one expert; it’s to create a process in which biased conclusions are more likely to be exposed.
Ironically, AI itself might become a useful auditing tool. Imagine requiring major platforms to allow qualified independent auditors to test their recommendation systems using thousands of simulated users with different browsing histories. Those auditors wouldn’t need access to the companies’ proprietary algorithms; they could evaluate the outputs statistically.
Now let me turn the tables for a moment.
Knowing you, I suspect your question wasn’t really about who writes the check.
I think you’re asking something closer to:
“Can we realistically expect an objective answer when the organizations with the most information and the most money may also have the strongest incentives to shape the narrative?”
That’s a much harder question.
My answer is that objective research is possible, but it requires transparency, replication, and investigators with different viewpoints. In science, we don’t usually trust one study. We trust findings that survive repeated scrutiny by people who would be delighted to prove the first researchers wrong.
And here’s one thing that might surprise you. I think there is actually one group with a strong incentive to conduct such research well: the platforms themselves. If a company could credibly demonstrate, through independently replicated studies, that its news recommendations were not systematically favoring one political viewpoint over another, that would be valuable to the company as well as to the public. The challenge is credibility. To earn public trust, the company would have to let genuinely independent researchers inspect and test the system rather than simply publishing its own conclusions.