[continued from previous message]

Welcome to the Artificial Intelligence Incident Database
Search over 3000 reports of AI harms
https://incidentdatabase.ai/

------------------------------

Date: Tue, 7 Jan 2025 14:32:38 -0800
From: "Jim" <jgeissman@socal.rr.com>
Subject: Apple's AI News Summaries and Inventions (BBC)

https://www.bbc.com/news/articles/cge93de21n0o

Apple is facing fresh calls to withdraw its controversial artificial intelligence (AI) feature that has generated inaccurate news alerts on its latest iPhones.

The product is meant to summarise breaking news notifications but has in
some instances invented entirely false claims.

The BBC first complained to the tech giant about its journalism being misrepresented in December but Apple did not respond until Monday this week, when it said it was working to clarify that summaries were AI-generated.


Alan Rusbridger, the former editor of the Guardian, told the BBC Apple
needed to go further and pull a product he said was "clearly not ready."

Mr Rusbridger, who also sits on Meta's Oversight Board that reviews appeals
of the company's content moderation decisions, added the technology was "out
of control" and posed a considerable misinformation risk.

"Trust in news is low enough already without giant American corporations
coming in and using it as a kind of test product," he told the Today
programme, on BBC Radio Four.

The National Union of Journalists (NUJ), one of the world's largest unions
for journalists, said Apple "must act swiftly" and remove Apple Intelligence
to avoid misinforming the public - echoing prior calls by journalism body Reporters Without Borders <https://www.bbc.co.uk/news/articles/cx2v778x85yo> (RSF).

"At a time where access to accurate reporting has never been more important, the public must not be placed in a position of second-guessing the accuracy
of news they receive," said Laura Davison, NUJ general secretary.

The RSF also said Apple's intervention was insufficient, and has repeated
its demand that the product is taken off-line.


Series of errors


The BBC complained <https://www.bbc.co.uk/news/articles/cd0elzk24dno> last month after an AI-generated summary of its headline falsely told some
readers that Luigi Mangione, the man accused of killing UnitedHealthcare CEO Brian Thompson, had shot himself.

On Friday, Apple's AI inaccurately summarised BBC app notifications to claim that Luke Littler had won the PDC World Darts Championship <https://www.bbc.co.uk/news/articles/cx27zwp7jpxo> hours before it began -
and that the Spanish tennis star Rafael Nadal had come out as gay.

This marks the first time Apple has formally responded to the concerns
voiced by the BBC about the errors, which appear as if they are coming from within the organisation's app.

------------------------------

Date: Tue, 31 Dec 2024 07:29:00 -0800
From: Lauren Weinstein <lauren@vortex.com>
Subject: What real people think about Google Search today

It's both notable and deeply depressing how many nontechnical people I know
who have unprompted told me how much they despise Google AI Overviews, which they inevitably describe as usually inaccurate and worthless, at which point they usually add how Google Search quality has declined enormously (in their own words, of course).

Then they sometimes say something like, "Hey Lauren, don't you know people
at Google that you could tell about how bad this is getting?"

At which point I usually bite my tongue, which is increasingly feeling like
a pincushion as a result.

Don't believe the happy face metrics that Google claims -- out in

------------------------------

Date: Fri, 10 Jan 2025 10:50:22 -0800
From: Lauren Weinstein <lauren@vortex.com>
Subject: WARNING: Google Voice is flagging LEGITIMATE robocalls from
insurance companies to their customers in the fires as spam

BE SURE TO CHECK YOUR SPAM FOLDERS! GOOGLE AI DOES IT AGAIN!

------------------------------

Date: Tue, 31 Dec 2024 10:28:03 -0800
From: Lauren Weinstein <lauren@vortex.com>
Subject: A non-tech analogy for Google Search AI Overviews

Here's a non-tech analogy to the problem (well, a problem) with Google AI Overviews:

Let's say you go to a restaurant. Maybe they're offering free meals
that day, maybe you're paying. Either way, several plates of
reasonable appearing food are placed in front of you. You ask about
the ingredients, but you only get vague answers back if any, and the
restaurant refuses to tell you anything about the actual recipes per
se.

You notice a little card sticking out from under one of the plates. It
reads:

"Some or all of this food may be fine. Some or all of this food may
have a bad taste. Some or all may give you food poisoning. It's up to
you to double check this food before eating it -- we take no
responsibility for any ill effects it may have on you."

Still hungry?

------------------------------

Date: Fri, 3 Jan 2025 09:58:24 -0500
From: Tom Van Vleck <thvv@multicians.org>
Subject: Happy new year, compute carefully

Just some notes to remind you to compute carefully in 2025.

1. In the past I recommended Gmail to people because it does some spam detection, but now Gmail is being exploited to hack people. If you get a (fake) call ostensibly from Google or (fake) notices that your Google
account is being attacked, run. Don't click anything. https://www.forbes.com/sites/zakdoffman/2025/01/03/new-gmail-outlook-apple-mail-warning-2025-hacking-nightmare-is-coming-true/?

2. If anybody says "now with AI," run.
They are not giving you something wonderful for free.

3. I have stopped using Google Chrome except for testing web page changes.
I avoid "Chrome Browser Extensions" because they have been hacked to do bad things.

4. 2.6 million devices have been backdoored with credential stealing
malware. Don't be a victim. https://therecord.media/hackers-target-vpn-ai-extensions-google-chrome-malicious-updates

------------------------------

Date: Sat, 4 Jan 2025 10:08:35 -0800
From: Lauren Weinstein <lauren@vortex.com>
Subject: How to understand Generative AI

To really understand generative AI, you need to keep one simple fact in
mind. There is no "Intelligence" in "Artificial Intelligence". OpenAI -- it turns out -- literally defines intelligence in terms of profits!

And as we see, Google AI is essentially a low grade moron. But this is true
for all of these systems. This is FUNDAMENTAL to how these systems
work. They are NOT intelligent. They do NOT understand what they're saying.

The term "Intelligence" in the context of these systems is merely a
MARKETING HYPE term, nothing more.

Keep this in mind and the chaos being created by Big Tech at our
expense is much easier to at least understand. -L

------------------------------

Date: Sat, 4 Jan 2025 16:51:29 -0800
From: Lauren Weinstein <lauren@vortex.com>
Subject: Google censoring my AI criticism?

One of the digest versions of today's mailings, which included
the messages:

1. The laughs keep rolling in to that fraction question I asked
Google (Lauren Weinstein)
2. The execs know their AI is trash (Lauren Weinstein)
3. Sources: Pentagon planning for how to deal with rogue Trump
(Lauren Weinstein)

was marked by Gmail as dangerous spam, with a red banner declaring it to
be a likely phishing attack. If you can figure out any possible way any
of those messages -- which were sent out as individual messages earlier
today -- could possibly be legit interpreted in that way, I'd love to
hear about it.

Otherwise, I suspect Google has filters in place to try divert some of
this criticism into a scary category that people won't read, whether
that was their actual intention or not.

VERY BAD. -L

------------------------------

Date: Sun, 5 Jan 2025 06:32:54 -0800
From: Steve Bacher <sebmb1@verizon.net>
Subject: U.S. newspapers are deleting old crime stories offering
subjects a clean slate (The Guardian)

Civil rights advocates across the US have long fought to free people from
their criminal records, with campaigns to expunge old cases and keep
people’s past arrests private when they apply for jobs and housing.

The efforts are critical, as more than 70 million Americans have prior convictions or arrests – roughly one in three adults. But the policies haven’t addressed one of the most damaging ways past run-ins with police can derail people’s lives: old media coverage.

Some newsrooms are working to fill that gap.

A handful of local newspapers across the US have in recent years launched programs to review their archives and consider requests to remove names or delete old stories to protect the privacy of subjects involved in minor
crimes.

“In the old days, you put a story in the newspaper and it quickly, if not immediately, receded into memory,” said Chris Quinn, editor of Cleveland.com and the Plain Dealer newspaper. “But because of our [search engine] power, anything we write now about somebody is always front and center.” [...]

https://www.theguardian.com/us-news/2025/jan/04/newspaper-crime-stories

------------------------------

Date: Thu, 9 Jan 2025 10:43:21 -0700
From: geoff goodfellow <geoff@iconia.com>
Subject: EU Commission Fined for Transferring User Data
to Meta in Violation of Privacy Laws (THN)

The European General Court on Wednesday fined the European Commission, the primary executive arm of the European Union responsible for proposing and enforcing laws for member states, for violating the bloc's own data privacy regulations.

The development marks the first time the Commission has been held liable
for infringing stringent data protection laws in the region.

The court determined that a "sufficiently serious breach" was committed by transferring a German citizen's personal data, including their IP address
and web browser metadata, to Meta's servers in the United States when
visiting the now-inactive futureu.europa[.]eu website in March 2022.

The individual registered for one of the events on the site by using the Commission's login service, which included an option to sign in using a Facebook account.

"By means of the 'Sign in with Facebook' hyperlink displayed on the E.U.
Login webpage, the Commission created the conditions for transmission of
the IP address of the individual concerned to the U.S. undertaking Meta Platforms," the Court of Justice of the European Union said in a press statement.

The applicant had alleged that by transferring their information to the
U.S., there arose a risk of their personal data being accessed by the U.S. security and intelligence services. [...] https://thehackernews.com/2025/01/eu-commission-fined-for-transferring.html

------------------------------

Date: Thu, 2 Jan 2025 09:22:06 -0800
From: Steve Bacher <sebmb1@verizon.net>
Subject: The Ghosts in the Spotify Machine (Liz Pelly:)

I first heard about ghost artists in the summer of 2017. At the time, I was
new to the music-streaming beat. I had been researching the influence of
major labels on Spotify playlists since the previous year, and my first
report had just been published. Within a few days, the owner of an
independent record label in New York dropped me a line to let me know about
a mysterious phenomenon that was “in the air” and of growing concern to those in the indie music scene: Spotify, the rumor had it, was filling its
most popular playlists with stock music attributed to pseudonymous musicians—variously called ghost or fake artists—presumably in an effort to reduce its royalty payouts. Some even speculated that Spotify might be
making the tracks itself. At a time when playlists created by the company
were becoming crucial sources of revenue for independent artists and labels, this was a troubling allegation. [...]

https://harpers.org/archive/2025/01/the-ghosts-in-the-machine-liz-pelly-spotify-musicians/

------------------------------

Date: Mon, 16 Dec 2024 09:35:13 -0800
From: Rob Slade <rslade@gmail.com>
Subject: Spotify

I have mentioned, at times, that many people seem to be laboring under the misapprehension that the email address rslade@gmail.com is theirs.

Recently I have had cause to look into Spotify. I don't carry my "tunes" around with me (well, they often pop up as mindworms, but I don't need any external source for that.), and I don't listen to podcasts, so I haven't
used Spotify, and I haven't created an account on it. But I've started contributing to a podcast, I didn't need to get a Spotify account to
listen to the podcast. But recently someone sent me a playlist of songs,
and I thought it would listen to it and hear what was in it. But Spotify, while it *would* play a free podcast, apparently *won't* play a playlist of commercial songs unless you create an account.

So I tried, only to find out, yes, you guessed it, there already *was* an account under the email address rslade@gmail.com. Of course, I didn't know
the account password. So, I just told Spotify that I lost the password.
And it helpfully sent me an opportunity to change it.

Whoever signed up for Spotify under my email address doesn't seem to have
any playlists or anything else on the account, so I guess they haven't used
it much and haven't lost anything. Much. Except for the account.

Handy for me, though ...

------------------------------

Date: Sat, 28 Oct 2023 11:11:11 -0800
From: RISKS-request@csl.sri.com
Subject: Abridged info on RISKS (comp.risks)

The ACM RISKS Forum is a MODERATED digest. Its Usenet manifestation is
comp.risks, the feed for which is donated by panix.com as of June 2011.

SUBSCRIPTIONS: The mailman Web interface can be used directly to

subscribe and unsubscribe:
http://mls.csl.sri.com/mailman/listinfo/risks

SUBMISSIONS: to risks@CSL.sri.com with meaningful SUBJECT: line that

includes the string `notsp'. Otherwise your message may not be read.
*** This attention-string has never changed, but might if spammers use it.

SPAM challenge-responses will not be honored. Instead, use an alternative

address from which you never send mail where the address becomes public!

The complete INFO file (submissions, default disclaimers, archive sites,

copyright policy, etc.) has moved to the ftp.sri.com site:
<risksinfo.html>.
*** Contributors are assumed to have read the full info file for guidelines!

OFFICIAL ARCHIVES: http://www.risks.org takes you to Lindsay Marshall's

delightfully searchable html archive at newcastle:
http://catless.ncl.ac.uk/Risks/VL.IS --> VoLume, ISsue.
Also, ftp://ftp.sri.com/risks for the current volume/previous directories
or ftp://ftp.sri.com/VL/risks-VL.IS for previous VoLume
If none of those work for you, the most recent issue is always at
http://www.csl.sri.com/users/risko/risks.txt, and index at /risks-34.00
ALTERNATIVE ARCHIVES: http://seclists.org/risks/ (only since mid-2001)
*** NOTE: If a cited URL fails, we do not try to update them. Try
browsing on the keywords in the subject line or cited article leads.
Apologies for what Office365 and SafeLinks may have done to URLs.

Special Offer to Join ACM for readers of the ACM RISKS Forum:

<http://www.acm.org/joinacm1>

------------------------------

End of RISKS-FORUM Digest 34.52
************************

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)