CNET Demoted to Untrusted Sources by Wikipedia Editors Due to AI-Generated Content

[AleksandarK]

Once trusted as the staple of technology journalism, the website CNET has been publically demoted to Untrusted Sources on Wikipedia. CNET has faced public criticism since late 2022 for publishing AI-generated articles without disclosing humans did not write them. This practice has culminated in CNET being demoted from Trusted to Untrusted Sources on Wikipedia, following extensive debates between Wikipedia editors. CNET's reputation first declined in 2020 when it was acquired by publisher Red Ventures, who appeared to prioritize advertising and SEO traffic over editorial standards. However, the AI content scandal accelerated CNET's fall from grace. After discovering the AI-written articles, Wikipedia editors argued that CNET should be removed entirely as a reliable source, citing Red Ventures' pattern of misinformation.

One editor called for targeting Red Ventures as "a spam network." AI-generated content poses familiar challenges to spam bots - machine-created text that is frequently low quality or inaccurate. However, CNET claims it has stopped publishing AI content. This controversy highlights rising concerns about AI-generated text online. Using AI-generated stories might seem interesting as it lowers the publishing time; however, these stories usually rank low in the Google search index, as the engine detects and penalizes AI-generated content probably because Google's AI detection algorithms used the same training datasets as models used to write the text. Lawsuits like The New York Times v. OpenAI also allege AIs must scrape vast amounts of text without permission. As AI capabilities advance, maintaining information quality on the web will require increased diligence. But demoting once-reputable sites like CNET as trusted sources when they disregard ethics and quality control helps set a necessary precedent. Below, you can see the Wikipedia table about CNET.



View at TechPowerUp Main Site | Source

 

[Psyclown]

cnet has mostly been a garbage tech site for quite a while now anyway.

 

[Chaitanya]

Its not like Wikipedia is a sqeaky clean when it comes to "information" hosted on its platform.

 

[the54thvoid]

Its not like Wikipedia is a sqeaky clean when it comes to "information" hosted on its platform.

Squeaky clean isn't a universal benchmark. Besides, anybody seeking knowledge should seek multiple sources. Wiki is still a suitable source. You dont like it? then find another source and compare. Hell, find many sources and draw a balanced conclusion. Of course, there are some who will patently refuse to see multiple viewpoints, but in that sense, Wiki is still more balanced than others, as it will tend to give opposing views in the same article, and/or, reference 'citation needed.' It's also an open format and therefore liable to bad actors. Just like any editorial press journal is open to bias.

 

[Kenjiro]

Its not like Wikipedia is a sqeaky clean when it comes to "information" hosted on its platform.

Yet people with names stand behind these and You can blame them. You can even correct articles if you are perfectly sure they are wrong.

 

[the54thvoid]

I fear this will get out of hand with a wiki attack - which is not the topic. The topic is CNET, use of AI , and it's subsequent demotion.

It's not really about what people think about wikipedia.

 

[Chaitanya]

I fear this will get out of hand with a wiki attack - which is not the topic. The topic is CNET, use of AI , and it's subsequent demotion.

It's not really about what people think about wikipedia.

Wikipedia also needs to be discussed as its they who have decided to demote CNET unilateraly on their platform. There are multiple "news" outlets who have started using machine learning tools recently to generate "articles" none of which seem to be flagged. I remember there was an article on Verge about the horrible state of new internet post ML tools and given nearly half of the world is heading into elections its not going to end well as with every new technology, its regulation playing catch up.

Squeaky clean isn't a universal benchmark. Besides, anybody seeking knowledge should seek multiple sources. Wiki is still a suitable source. You dont like it? then find another source and compare. Hell, find many sources and draw a balanced conclusion. Of course, there are some who will patently refuse to see multiple viewpoints, but in that sense, Wiki is still more balanced than others, as it will tend to give opposing views in the same article, and/or, reference 'citation needed.' It's also an open format and therefore liable to bad actors. Just like any editorial press journal is open to bias.

Wikipedia has a massive problem with deletion of information which doesnt fit the political biases of its editiors(just the tip of iceberg of issues with wiki foundation) and such citations and information are quick to get deleted. As such there was a court ruling regarding wikipedia being tainted and unreliable source of information and cannot be used for any legal matters.

 

[Fourstaff]

I wonder if its a wholesale demotion, or its a selective demotion from 2020 onwards. Really hoping Wikipedia can be a good judge of the incoming misinformation wave.

 

[windwhirl]

I wonder if its a wholesale demotion, or its a selective demotion from 2020 onwards. Really hoping Wikipedia can be a good judge of the incoming misinformation wave.

Selective depending on time period. For older articles, since CNET deleted or is deleting a lot of those, an Internet Archive link is recommended.

https://en.m.wikipedia.org/wiki/Wikipedia:Reliable_sources/Perennial_sources

 

[LazyGamer]

The demotion of CNET from Trusted to Untrusted Sources on Wikipedia underscores the significant impact of ethical lapses in journalism, particularly concerning AI-generated content. While the convenience of AI-written articles may seem appealing, the lack of transparency and potential for misinformation pose serious concerns for maintaining information integrity online. CNET's failure to disclose AI involvement and prioritize editorial standards has rightfully drawn criticism and led to its loss of credibility. This controversy serves as a stark reminder of the importance of ethical considerations and quality control in the era of advancing AI capabilities. Setting such precedents is crucial for upholding trustworthiness and reliability in digital media landscapes.


P.S.
This response was A.I. generated

 

[the54thvoid]

Wikipedia has a massive problem with deletion of information which doesnt fit the political biases of its editiors(just the tip of iceberg of issues with wiki foundation) and such citations and information are quick to get deleted. As such there was a court ruling regarding wikipedia being tainted and unreliable source of information and cannot be used for any legal matters.

Quoting an article over 5 years old containing research that dates back as far as 2011 isn't a great way to prove your point (in this discussion). As for a court ruling - ofc that would be the case. Press articles aren't legal documents. Wiki is an online open journal.

But it's good you've highlighted the major problem wiki had/has: that of male dominance. Second to that is the western slant. But, as it's a male-dominated westernised output source, then it seems fitting for it to comment on CNET, which is also male dominated western output. No?

Other than that, IMO, it's good that a press outlet has been red flagged for using AI as its journalistic output. I figure the best way an AI could serve journalism is for it to be used to make it grammatically sound and remove emotional bias from an author's article. In other words, you write about something and your personal bias seeps in but you ask AI to rewrite it and remove emotional tags. Make it more neutral as a source of info.

 

[R-T-B]

Its not like Wikipedia is a sqeaky clean when it comes to "information" hosted on its platform.

It's considerably clean when you consider the editors can be literally anybody online.

I wonder if its a wholesale demotion, or its a selective demotion from 2020 onwards. Really hoping Wikipedia can be a good judge of the incoming misinformation wave.

If it truly is, it'll get labeled as misinformation by some quickly, mark my words. Fake news genie can't be put back in the bottle.

 

[Tahagomizer]

It's weird that CNET was considered a trusted source at any point. In the best of times they were nothing more than a corporate shill, gushing over any rubbish produced by their advertisers.
As for the rest, LLMs are a wonderful tool when used for their intended purpose. Sadly, currently the Internet is just scripts scraping popular words, feeding them to an LLM and publishing the result with no vetting or proofreading. Finding any information in this pile of SEO excrement is increasingly difficult.

 

[ThrashZone]

Hi,
CNet well no better place to get malware downloads hehe

AI well wouldn't it be sourcing from wiki content seeing it's basically a search engine coded like google/ bing/... searches are pay to play or in search pay to be on top so pony up some AI bucks wiki

 

[xorbe]

Even motorcycle websites are using AI for content. These "articles" tend to be regurgitated trash. There's also a rash of "guided AI posts" in forums trying to drive/increase "user engagement".

 

[Double-Click]

Says a lot about CNET if it wasn't apparent already to most.
If you care at all about the integrity of your work as an author (or your media outlet stepping back a bit) you would never let AI just take the wheel.

This is in part why the writers guild wigged so hard about AI with their contracts.
Trying to use AI in place of actual human thought and originality is morally bankrupt.

 

[Chrispy_]

They completely died in 2020. Some scam artists bought the name and just cashed in on its prior reputation. This latest demotion to "untrusted" is just one of many milestones officially confirming CNET's rapid death after 2020's buyout.

This wouldn't be the first time a struggling news site was bought up by some VCs, staff reduced to a skeleton crew, and journalism replaced by AI content that's geared towards attracting SEO engines instead of any kind of quality journalism.

RIP SharkyExtreme, Anandtech, Techreport, and more.

 

[Heiro78]

Wikipedia also needs to be discussed as its they who have decided to demote CNET unilateraly on their platform. There are multiple "news" outlets who have started using machine learning tools recently to generate "articles" none of which seem to be flagged. I remember there was an article on Verge about the horrible state of new internet post ML tools and given nearly half of the world is heading into elections its not going to end well as with every new technology, its regulation playing catch up.


Wikipedia has a massive problem with deletion of information which doesnt fit the political biases of its editiors(just the tip of iceberg of issues with wiki foundation) and such citations and information are quick to get deleted. As such there was a court ruling regarding wikipedia being tainted and unreliable source of information and cannot be used for any legal matters.

Thanks for the link to the guardian article on the biases. Got one for that court ruling as well?

CNET does indeed suck though. Just recently I was looking to complain to them about their promotion of a scam lifetime 2+TB cloud storage deal. Users had been complaining that the service would vastly limit the amount of storage, their access to the files, or incredibly slow speeds. It sucks since I remember them being a great resource for tech news. Atleast we still got TPU!

Hi,
CNet well no better place to get malware downloads hehe

AI well wouldn't it be sourcing from wiki content seeing it's basically a search engine coded like google/ bing/... searches are pay to play or in search pay to be on top so pony up some AI bucks wiki

I can confidently say, this was not written by ai....

 

[ThrashZone]

I can confidently say, this was not written by ai....

Hi,
Yeah I know what I said was way to true

 

[Chaitanya]

Thanks for the link to the guardian article on the biases. Got one for that court ruling as well?

CNET does indeed suck though. Just recently I was looking to complain to them about their promotion of a scam lifetime 2+TB cloud storage deal. Users had been complaining that the service would vastly limit the amount of storage, their access to the files, or incredibly slow speeds. It sucks since I remember them being a great resource for tech news. Atleast we still got TPU!


I can confidently say, this was not written by ai....

Here is the link to Indian Supreme court judgement(behind paywall) about Wikipedia being unreliable source of information(it was Hp and Lenovo who had cited wikipedia leading to that ruling check pdf). I remember reading there are couple of other lawsuits against them right regarding violation of donations and their part in spreading misinformation.

'Wikipedia Not Completely Dependable' : Supreme Court Cautions Courts & Adjudicating Authorities

In a judgment delivered today, the Supreme Court cautioned the courts and adjudicating authorities against use of 'wikipedia' for legal dispute resolution.These sources, despite being a treasure...

www.livelaw.in

 

[R-T-B]

Hi,
CNet well no better place to get malware downloads hehe

AI well wouldn't it be sourcing from wiki content seeing it's basically a search engine coded like google/ bing/... searches are pay to play or in search pay to be on top so pony up some AI bucks wiki

I'm sorry but this is far from true, it doesn't even make sense as a coherent statement (wiki wouldn't be the one paying for searches, AI would, and it doesn't work like that anyways...)

 

[sLowEnd]

They're still around? I haven't used CNET since like...2007?

 

[ExcuseMeWtf]

Let's not poison the well.
I acknowledge there are tons of issues with Wikipedia itself, but this is clearly a correct course of action regarding CNET.

 

[Chrispy_]

Let's not poison the well.
I acknowledge there are tons of issues with Wikipedia itself, but this is clearly a correct course of action regarding CNET.

Wikipedia can be temporarily inaccurate because the number of contributors outnumber moderators. Eventually though, content that fails to pass factual or impartiality checks gets updated. It's kind of self-correcting because the minute someone cites an impartial/biased Wikipedia entry as evidence in an argument of any significance, the more incentive there is to flag that entry to moderators for review.

 

[remixedcat]

Even motorcycle websites are using AI for content. These "articles" tend to be regurgitated trash. There's also a rash of "guided AI posts" in forums trying to drive/increase "user engagement".

entire youtube channels that are made and managed with AI too.... ppl that then gloat about thier AI channels making 50k/mo too