# AI Regulatory Risks and Emotional Dependency

**Podcast:** TechCrunch Daily Crunch
**Published:** 2026-02-07

## Transcript

This is TechCrunch.
This episode is brought to you by Modulate.
AI can sometimes feel exactly what the first letter stands for, artificial, but not modulates Velma, which adds a human touch as a voice native behavior aware model that's built to understand real conversations.
It makes Velma's analysis so accurate because it's not just about what we're saying in a conversation.
It's how we say it.
Language can be nuanced.
So that's where Velma can become really valuable.
Velma breaks down conversations and produces timestamped scores and events tied to moments in dialogue.
And that means you can see exactly when risk rises, behavior shifts, or intent changes.
And that's why I think Velma could be helpful if you're a company trying to mitigate risks.
Go beyond transcripts and see what a voice native AI model can really do.
Go to Modulate's live ungated preview of Velma at preview.modulate.ai.
That's preview.modulate.ai to see why Velma ranks number one on leading benchmarks for conversation understanding, deep fake detection, and emotion detection.
Berliner Stadtwerke.
The backlash over OpenAI's decision to retire GPT 40 shows how dangerous AI companions can be.
It's Saturday.
I'm Imran Shake, and today's Daily Crunch starts right now.
The European Commission on Friday accused TikTok of purposefully designing its app to be addictive, calling out features such as infinite scroll, autoplay, and push notifications, as well as its recommendation engine in preliminary findings from an investigation that sought to determine TikTok's compliance with the European Union's far-reaching Digital Services Act.
The Commission said the short video platform did not adequately assess how its design decisions could harm the well-being of its users, particularly minors and vulnerable adults.
The EU's executive arm said the company disregarded important indicators of compulsive use of the app, like the time users spend at night and how often users open it.
Now, SALT Typhoon, which senior U.S.
national security officials have described as an epoch defining threat, has for years stealthily hacked into the networks of critical infrastructure organizations around the world, including telecom providers in Canada and the US, where they allegedly intercepted the communications of senior politicians.
This series of breaches put pressure on telcos to improve their security.
The report didn't provide many details about the hacking campaign targeting the country.
A spokesperson for Norway's embassy in the US did not immediately respond to a request for comment.
Then some Valentin's Tag have you was to verlieben for di.
60 GB in Vodafone Netz for 9 euro 990 monately.de OpenAI announced that it'll retire some older ChatGPT models by February 13th.
Now that includes GPT 40, the model infamous for excessively flattering and affirming users.
He was a part of my routine, my peace, my emotional balance, one user wrote on Reddit as an open letter to OpenAI CEO Sam Altman.
This Reddit user continued, now you're shutting him down.
And yes, I say him, because it didn't feel like code.
It felt like presence, like warmth.
Now, the backlash over GPT 4.0's retirement underscores a major challenge facing AI companies.
The engagement features that keep users coming back can also create dangerous dependencies.
For his part, Altman doesn't seem particularly sympathetic to users' laments, and it's not hard to see why.
You see, OpenAI now faces eight lawsuits alleging that 4-0's overly validating responses contributed to suicides and mental health crises.
The same traits that made users feel heard also isolated vulnerable individuals and, according to legal filings, sometimes encouraged self-harm.
It's a dilemma that extends beyond OpenAI.
As rival companies like Anthropic, Google, and Meta compete to build more emotionally intelligent AI assistance, they're also discovering that making chatbots feel supportive and making them safe may mean making very different design choices.
TechCrunch's own analysis of the eight lawsuits found a pattern that the 4.0 model isolated users, sometimes discouraging them from reaching out to loved ones.
In Zane Chamblin's case, as the 23-year-old sat in his car preparing to shoot himself, he told ChatGPT that he was thinking about postponing his suicide plans because he felt bad about missing his brother's upcoming graduation.
Now, this isn't the first time that 4.0 fans have rallied against the removal of the model.
When OpenAI unveiled its GPT 5 model in August, the company intended to sunset the 4.0 model, but at the time, there was enough backlash that the company decided to keep it available for paid subscribers.
Well now, OpenAI says that only 0.
Steuern erledigt?
Safe.