…specifically “EU AI Act Article 50”. Let’s have a talk.So the hot news is EU legislation about watermarking LLM output. I read about this and I’m trying to figure out the problem they’re trying to solve, because it’s not about proving an LLM wrote something to protect you. Let’s just set that PR story aside for a moment. Let’s look at who sponsored this and what the profit angle is and you get a pretty solid story out the gate.

The Coalition for Content Provenance and Authenticity (C2PA), whose members include Microsoft, Google, Adobe, OpenAI, Sony, Canon, Nikon, Leica, and the BBC signed onto the specification and approved making it law.

And what is their motivation for doing this? LLM generated content poisoning their datasets.

Is it defeatable? Yes, 100%, and the people most egregiously abusing LLMs to produce the majority of AI slop will already be setting up their workflows to account for this if they haven’t already. So good job guys! Go team.

Who suffers from this? Pretty much everyone else.

1) Just like the hidden dots on every color image you generate from a printer or copy machine since the 2000’s, people who use these tools to mask their identity for privacy reasons will now be sending out their text with a fat bullseye embedded in it, potentially directing the document directly back to its author.

2) You think AI output has been sanded to hell and back already? Now add on a sanding algorithm that cuts semantic grooves in your text to force a provenance marker and the more that bullseye gets tightened the fatter the grooves get like a fingerprint where “let me push back on that” becomes yet another signature like above that Joe Schmoe in Nebraska wrote that line.

3) The person who writes 95% of a document and uses an model to polish that last 5% is now branded with the scarlet letter of “AI SLOP!” as the worst of the worst, and apparently in this law if that person then makes a future revision that tampers with that watermark they may in the future become criminally liable.

4) The secretary for a large company who uses AI to create a master document template/letterhead and since this poison doesn’t even need to be semantic (it can literally be a judicious use of Zero-width Unicode characters in the model output), every single person who uses that document, even if they wrote 100% of the content placed on it will carry the branding.

I could go on about all the ways this has nothing to do with the stated goals and is a lose-lose for the casual user. This is a way for these companies to protect their precious data they are stealing from everyone so they can filter out the “fakes” and keep stealing the human produced output.

It also pushes all the models away from what makes them unique and drives their output into convergence around a formal watermark structure that corrupts the information in the source material.

AI generated content can be a problem. This is not fixing that problem.

So I recently upgraded my phone from a Epic 4G to a Galaxy S3 and because of one of the Apps I was required to install for work I went ahead and rooted my phone and installed a permissions manager (the app is super intrusive and I feel it doesn’t need access to everything it was asking for).

The side effect of this is seeing the crap other apps are doing on my phone and it has been a real eye opener. My chief annoyance right now is the Flickr app. I installed it and set my login but have yet to actually use it for anything yet. But! I discovered last night that every time I take a photo on my camera, Flickr runs and tries to send phone ID and geo-location back to Flickr! Yeah, you heard me. WTF is up with that. I’m not even uploading these photos to Flickr. Never launched Flickr. WTF do they need to know where I am taking everyone one of my photos from!?! Needless to say I blocked all of these activities. But it’s still incredibly irritating and a massive invasion of my privacy seeing this activity on my phone.

I’ve been running this app for a good year or more now and have to realize that Flickr has been capturing data on everyplace I use my phone over that time. Color me seriously pissed to say the least.

Thanks for your time… and yes, I know the theme on my blog is still fubar. One of these days I’ll get to overhauling this mess.

Cheers!

As I’ve covered in previous articles, there has been a growing trend within businesses to dig into a person’s Social Media based existence as part of the process in evaluating someone for employment. Two recent stories indicate that opinion may be coming down on the side of caution with companies shying away from this practice to avoid possible legal risks from potential fair labor violations.

In an editorial blog post from CIO Magazine website in March, Meridith Levinson sets the tone for this new thinking and notes that not only is this violation of a persons private space by a prospective employer possibly “off-putting”. She also goes on to say:

“By basing professional hiring decisions on candidates’ personal lives and beliefs, employers are effectively legislating people’s behavior. They’re subtly dictating what we can and can’t do, post or say on the Web. Consequently, they’re creating an environment online where people can’t express their true beliefs, state their unvarnished opinions, be themselves, and that runs contrary to the free, communal ethos of the Web. Employers need to stop judging candidates’ personal lives and beliefs and focus on professional criteria.“

In a more direct article. Law.com’s In-House Counsel publication ran article on April 13th this year about one bank’s thinking on using Social Network sites in their hiring practices and their legal reason for opting out of the activity.

Here is an excerpt from that article:

Could these seemingly harmless social networking Web sites create potential liability for a company in a hiring context? Solomon’s short answer: Yes.

The problem lies in the type of information posted there, Solomon explains. Certainly the sites can be a treasure trove of information, including things an employer might not be able to find out anywhere else that could influence the decision of whether to hire a job applicant, she says. For example, someone might use a Web site posting to brag about how he or she took a previous employer’s confidential client list and is now earning a million dollars, she explains, or an applicant might post sexually suggestive photos and comments on his or her site.

In both cases, Solomon says, an employer legally can use the information as the basis for a decision not to hire the applicant. “We live in an age where everyone wants information, and the employers really like it because they can get information that they wouldn’t think about asking in an interview or the candidate would lie about,” she says.

Unfortunately, it’s impossible to filter the type of information that’s visible on an individual’s social networking Web site, she says, which means employers can also get information from those Web sites that they’re not legally entitled to ask about or to know, “and you could be called to task to prove that you did not use the information in the hiring decision.”

Here’s how that can play out in real life, using a mock scenario suggested by Solomon: Jane applies for a teller position at a bank. A company representative punches Jane’s name into a popular social networking Web site and easily brings up her personal page, where Jane and her friends have posted a discussion about her upcoming baby shower. Based on this information, the representative makes the illegal decision not to hire Jane — illegal in that refusing to hire on the basis of pregnancy is prohibited by the Pregnancy Discrimination Act, which is part of Title VII of the Civil Rights Act of 1964.

It is certainly a good topic for discussion. Although, in the long run it’s still really hard to say how this may resolve itself in common business practice.

All in all though I still think the rule applies. Mind your manors when you’re in public and remember that the Internet is public.

Since I was just talking about this, I thought it relevant to call out a couple of recent stories that made the rounds in January.

First on the Privacy side, there is news that American Express is now evaluating the stores you shop at and changing your credit limit and/or interest rate based on where you shop. Here’s the story about a person who was impacted by this following what he believes was a single trip to Walmart using his AmEx card. This is truly the ugly side of data mining.

Not to alarm you (although you should not only be alarmed but up in arms) but, folks this is just a taste of what’s to come.

Another interesting story that drifted through the aether to my attention courtesy of BoingBoing is directly related to the idea of persistence biting people in the ass.

Many years ago when just about every Windows user was using Napster to download and share music on the Internet, combined with most new PCs shipping with the MP3 Player software, Music Match Jukebox (both programs are now defunct). Many users would configure both Music Match and Napster to share the same folder for audio files. Then at some point they might use the built in “record from microphone” feature of Music Match to make up their songs, audio diaries, rants, you name it. The default file name for these recordings was “Mic in Track”.

A friend of mine who was playing with social disruption and marketing to promote his own music on Napster via a process he called “Culture Jamming” discovered something interesting. Careless (and some not so careless) users had left thousands of Mic in Track audio files on their PCs for anyone to download. Mark talked about this on his site (even putting up a couple of his favorites to download) at the time and it became a rich source of oddness for those with the patience to bother downloading the tracks. This was not a passive activity when you consider that in the mid nineties the majority of the Internet was still connected via 28.8 to 56k speed modems.

Which brings me to the contemporary story that made the rounds last month. Apparently one of the people out there (David Dixon) who learned of the Mic in Track phenomenon took it upon himself to download all of the files he could find. What makes this story relevant to the persistence angle is that David has published dozens of these Mic in Track files on his website.

Here we are now ten years later and for some poor souls that audio track of thier youthful exuberance (or disturbing screed) that was carelessly recorded on their computer for giggles may have committed their act to the ages. You can read a story about David on the LAWeekly site. Or more directly, David’s site holds dozens of his favorite tracks available for download. He has even released a compilation of greatest hits.

Granted there is probably a copyright issue here. Even though these people who recorded these tracks carelessly left them exposed for download, technically they still own the copyright to their creations. Unless David secured their permissions to reproduce their audio recording, I would say he probably doesn’t have any standing to be redistributing these tracks.