Tech
OpenAI releases a new safety blueprint to address the rise in child sexual exploitation
In response to escalating concerns about child safety online, OpenAI has unveiled a blueprint to enhance U.S. child protection efforts amid the AI boom. The Child Safety Blueprint, which was released Tuesday, is designed to help with faster detection, better reporting, and more efficient investigation into cases of AI-enabled child exploitation.
The overall goal of the Child Safety Blueprint is to tackle the alarming rise in child sexual exploitation linked to advancements in AI. According to the Internet Watch Foundation (IWF), more than 8,000 reports of AI-generated child sexual abuse content were detected in the first half of 2025, a 14% increase from the year prior. This includes criminals using AI tools to generate fake explicit images of children for financial sextortion and to generate convincing messages for grooming.
OpenAI’s blueprint also comes amid increased scrutiny from policymakers, educators, and child-safety advocates, especially in light of troubling incidents where young individuals died by suicide after allegedly engaging with AI chatbots.
Last November, the Social Media Victims Law Center and the Tech Justice Law Project filed seven lawsuits in California state courts, alleging that OpenAI released GPT-4o before it was ready. The suits claim the product’s psychologically manipulative nature contributed to wrongful deaths by suicide and assisted suicide. They cite four individuals who died by suicide and three others who experienced severe, life-threatening delusions after extended interactions with the chatbot.
This blueprint was developed in collaboration with the National Center for Missing and Exploited Children (NCMEC) and the Attorney General Alliance, as well as with feedback from North Carolina Attorney General Jeff Jackson and Utah Attorney General Derek Brown.
The company says that the blueprint focuses on three aspects: updating legislation to include AI-generated abuse material, refining reporting mechanisms to law enforcement, and integrating preventative safeguards directly into AI systems. By doing so, OpenAI aims not only to detect potential threats earlier but also to ensure actionable information reaches investigators promptly.
OpenAI’s new child safety blueprint builds on previous initiatives, including updated guidelines for interactions with users under 18, which prohibits the generation of inappropriate content, or encouraging self-harm, and avoiding advice that would help young people conceal unsafe behavior from caregivers. The company recently released a safety blueprint for teens in India.
Techcrunch event
San Francisco, CA
|
October 13-15, 2026
Tech
Hacker stole £700,000 from UK energy company by redirecting payment
British oil and gas company Zephyr Energy says someone stole £700,000 (close to $1 million) from one of its U.S.-based subsidiaries by redirecting a payment meant for a contractor into a hacker-controlled account.
In a regulatory filing with the London Stock Exchange on Thursday, the company said it is “working with the corresponding banks and consultants to attempt to recover the diverted funds.”
While the company did not say how the incident occurred, hackers are known to break into email inboxes or accounting systems and use that access to alter bank account and routing numbers during the process of paying someone or clearing an invoice. Known as business email compromise attacks, the FBI said in its most recent annual report published on internet cybercrime earlier in April that these attacks remain one of the top sources of financial losses, totaling more than $3 billion in victim losses during 2025.
Zephyr says that its incident is contained and that its operations are running normally.
As for the attack itself, the company said it used “industry standard practices” for its tech and payment platforms, but said it has implemented “additional layers of security” following the incident.
A spokesperson for Zephyr did not return an email requesting comment about the incident.
(via The Register)
Tech
X brings back Voice Notes to X Chat
Posting Voice Notes publicly on X may no longer be possible, but you can now share audio messages within X’s direct messaging system, X Chat, once again. The social network announced late on Wednesday that support for Voice Notes is now available within its private messaging service.
The feature, which works in both one-on-one messages and group chats, is activated with a push of the voice input icon to the right of the chat’s text box. At launch, you have to continue to press the button to record the voice message, but we found that a press-and-hold gesture followed by a swipe up allows it to record without having to keep your finger on the button.
The new addition could make X Chat more competitive with other messaging apps, where recording audio voice notes has long been a standard option. This is particularly important to the company, given the recent spinout of X Chat as its own stand-alone app.
It could also assuage angry users who didn’t appreciate that the upgrade to X Chat removed the Voice Notes feature.
The move follows X’s recent beta tests of an X Chat app on iOS, which offers access to X’s upgraded DM feature. While the company claims that chats are end-to-end encrypted, security experts have warned that the service is less secure than other encrypted messaging apps, like Signal.
The introduction of the new app reflects a strategy change for the social network, as owner Elon Musk once said X would become an all-in-one super app, or “everything app.” Now, the company is looking to make pieces of its app available as their own experiences. X Money, X’s payments service, is also being tested as a separate app, for instance.
Voice Notes have been on X Chat’s roadmap for some time despite their temporary removal. When X first introduced its new chat platform in November, it said the audio feature would be “returning soon.”
Currently, the X Chat service also supports other features, like the ability to edit and delete messages, block or get notified of screenshots, share files, make voice and video calls, and set messages to automatically disappear.
Tech
Avec’s Tinder-style email app allows you to swipe through your inbox
Apps like Superhuman and Mimestream have tried to get people to inbox zero on the desktop. Now a new app called Avec for mobile devices, initially available on iOS, aims to get you through your inbox using Tinder-style swipe cards and voice-based replies.
By default, the left swipe adds the email to a pile that you can address later, and the right swipe adds it to the done (or archive) pile.
The email “stack” of cards also has a button at the bottom that lets you hold it to reply to emails using your voice. When you release the button after speaking, the transcription will show up as a draft. You can review the transcription for errors, make any necessary edits, and then send the email.
Avec said that while apps like Wispr Flow, Willow, and Monologue exist, they are constrained by Apple’s APIs, and users need to install them as a separate keyboard app to work. Meanwhile, Avec has the full context of your email, so it can understand names and apply better edits based on the tone of the email. Because of this context, the email app can understand your personal email style as well, the company said.

While managing your inbox, Avec lets you mark unimportant emails by swiping down. The email will learn from what’s put in the unimportant pile and can show it to you in a group instead of forcing you to triage these emails one by one.
While the card-based interface is Avec’s unique feature, it also offers a plain old list-based view.
The app was founded by Jonathan Unikowski, who previously worked at Replit in a product engineering role. Unikowski said he was thinking about building tools that he would use every day. He explored ideas like building a browser but eventually ended up with email.
Techcrunch event
San Francisco, CA
|
October 13-15, 2026
“It’s this thing that hasn’t changed for 25 years,” Unikowski told TechCrunch over a call. He said Gmail was the last big change in email, which has had long-term impacts on how email is managed. “It’s a big part of everyone’s life, no matter how much they hate it. And it seemed very clear to me that through a combination of really good design and, of course, the judicious use of these new AI tools, we could do much better.”

Avec is not alone in having this thought process. Apart from Superhuman, apps like Shortwave and Spike have tried different approaches to presenting email. In the last decade, Basecamp’s Hey has tried to “reinvent” email by becoming a new provider, but, as a paid service, it hasn’t reached the same scale as Gmail.
When I asked Unikowski about choosing mobile over desktop as a first place to launch an email client, he said that constraints on the platform can force creativity, and the phone is usually the place where people look at their emails.
“I really am a firm believer in this idea that constraints force creativity, and so you get away with a lot less on an iOS app. On phones, you have a very small screen [as compared to the desktop]. You don’t have a physical keyboard. So if you’re going to convince someone to install a new app, it needs to be really good. And for it to be really good, you need to be extremely inventive,” he said.
The app is currently available in the U.S. and is free to use for Gmail users. Support for Outlook is in the works. Unikowski said that the company plans to introduce paid tiers at some point, but it is still ideating about what features to include within that premium offering.
The company has raised $8.4 million in funding to date from investors, including Lightspeed and Haystack, with participation from individuals such as Replit CEO Amjad Masad, Replit’s head of AI Michele Catasta, Behance co-founder Scott Belsky, and Lenny Rachitsky.
