Technology

When Meta embedded AI bots in its apps, even students in the most remote corners of Colombia gained access. But rather than boosting learning, it’s getting in the way.

• Meta’s popularity has made it the first place many Colombian students encounter and use generative AI.
• Teachers report a surge in AI-generated homework and essays, while student performance on exams declines.
• Educators warn easy access to AI is deepening existing problems in the country, which was already struggling with low graduation and literacy rates.

I found the real Spotify accounts of celebrities, politicians, and journalists. Many use their real names. With a little sleuthing, I could say with near-certainty: yep, this is them.

We’ve been scraping their accounts since summer 2024. Playlists, live listening feed, everything. I know what songs they played, when, and how many times.

Heard of the Panama Papers? That exposed offshore bank accounts. This is about onshore vibes.

Commit.

Source.

YouTube announced July 29, 2025, that videos containing strong profanity in the first 7 seconds will now be eligible for full ad revenue, reversing a policy that previously limited monetization for such content. The update, revealed through Jensen from TeamYouTube, eliminates restrictions that caused videos to earn limited or no ad revenue when strong language appeared during the opening moments.

According to the announcement, the policy originally aligned YouTube's content standards with television broadcasting guidelines. "The policy originally aligned YouTube's content standards with the guidelines set for TV," the update states. However, advertiser tools now provide businesses with choices over ad placement based on content profanity levels, reducing the need for blanket restrictions.

our results indicate that hate speech communities on Reddit share speech patterns with Reddit communities for Schizoid Personality Disorder, three Cluster B personality disorders (Borderline, Narcissistic, and Antisocial), and Complex Post-Traumatic Stress Disorder. While the Cluster B disorders would be expected due to prior studies into the Dark Triad and hate speech, this nonetheless offers confirmation that they are acting similarly to their Dark Triad counterparts when it comes to hate speech. Furthermore, the two non-Cluster B disorders have not been discussed in hate speech or misinformation literature to the best of our knowledge, and offer new routes of investigation.

The association in speech patterns between certain psychiatric disorders and hate speech (and misinformation to a lesser extent) suggests that, despite hate speech and misinformation being a social phenomenon, framing and approaching hate speech/misinformation as though they were psychiatric disorders may prove beneficial in combating them. For example, designing counter-messaging against these issues could utilize elements of therapy used for the psychiatric disorders that are most similar to hate speech and misinformation. Although hate speech embeddings being classified most frequently as Cluster B personality disorders could be seen as problematic due to those disorders’ historical resistance to treatment, the successes in treating Borderline Personality Disorder through Dialectical Behavior Therapy indicates that treatment of these disorders is not impossible and could be applied here [59,60]. Furthermore, although hate speech shares similarities with Cluster B personality disorders, that does not mean that it actually is a Cluster B personality disorder; it is most likely less resistant to change than a personality disorder is.

Misinformation’s relations to psychiatric disorder communities proved to remain elusive; a clear connection can be seen in the TDA mapping, and zero-shot classification had over 25% of the misinformation embeddings assigned to psychiatric disorders, but it was difficult to elucidate a clear pattern with the dataset at hand. A more robust dataset may be able to identify clearer trends.

Cheat Sheet Source

• Main Thread.
• Second Thread.
• Another Thread.

Strategy in Practice:

Original article from Hypertext, republished under Creative Commons Attribution-NonCommercial-ShareAlike 4.

• Tea, a dating safety app for women, is the subject of an incredibly alarming data breach.
• Tens of thousands of images submitted by users, including selfies, have been ripped from a Tea server and were posted to 4chan before being removed.
• Despite claiming that the breach only affected users who registered before February 2024, it has now come to light that hackers could read DMs between users as recently as a few weeks ago.

No business wants to shout from the rooftops that it has been breached and that the data their users entrusted them with may be circulating on the internet. It’s bad for public relations and destroys trust. However, just because it feels bad doesn’t mean that the custodians of this data can just sweep a breach under the rug.

Case in point is Tea. Tea is a dating safety app where women can share information about their previous partners in a bid to help other women who may encounter these men in the wild. Tea takes the Facebook groups and cobbled-together websites of old and puts a modern, more easily accessible twist on the practice.

Last week, however, the platform was the subject of a breach.

“We discovered unauthorised access to an archived data system,” Tea wrote in a post on its Instagram page.

“This archived system stored about 72 000 user-submitted images including approximately 13 000 images of selfies and selfies including photo identification submitted during account verification. These photos can in now way be linked to posts within Tea,” the developer wrote.

The company claimed that users who signed up for Tea after February 2024 were safe and that no email addresses or phone number were compromised. However, that’s ignoring the thousands of users who now have their data exposed. Worse still, that data system Tea mentions was posted to 4chan before it was eventually removed.

While photos can’t be linked to accounts, that’s besides the point because even just having one’s ID photo in the data dump could be incredibly dangerous for women.

And to make matters worse, somehow there has been a second incident.

As reported by [404 Media, a security researcher has discovered that it was possible for hackers to access messages between users as recently as last week. This flies in the face of Tea’s statement that no current user data is in danger. As the publication puts it, “it was trivial for 404 Media to find the real world identities of some users given the nature of their messages.”

All this while Tea continues to downplay how serious this is for its users.

Even the developer’s reasoning for why the data was breached is weak as it gets.

“During our early stages of development some legacy content was not migrated into our new fortified system. An unauthorized actor accessed our identifier link where data was stored before February 24, 2024. As we grew our community, we migrated to a more robust and secure solution which has rendered that any new users from February 2024 until now were not connected to the images involved in this incident,” Tea writes in an FAQ.

Excuse us, but what? There was an unsecured database just left somewhere in its system since last year, and Tea did nothing about it. That doesn’t sound like “dating safety tools that protect women” as the app proclaims on its website.

This should be grounds for a business-ending fine because, for the users, there is frankly nothing they can do. Their photos, possibly their messages, and more are now compromise,d and while the database containing that info was removed from 4chan, it could now be just about anywhere.

However, Tea’s social media posts about this breach are awash with users who are begging for Tea to accept their application to join the platform. One user even told the platform, “we don’t care about the leak” which is mighty concerning. There are some who are calling for Tea to rebuild and return with a safer app for the users, but the most vocal commenters simply want access.

What’s next for Tea? We honestly don’t know. A breach like this should be the end for a company, but it seems that Tea’s popularity has outweighed the danger of this incident and will likely grow as time marches on because, despite its security failings, there is a demand for this sort of thing.