Technologies

Claude AI Can Now End Conversations It Deems Harmful or Abusive

The feature marks a milestone in how Anthropic approaches AI safety.

Anthropic has announced a new experimental safety feature that allows its Claude Opus 4 and 4.1 artificial intelligence models to terminate conversations in rare, persistently harmful or abusive scenarios. The move reflects the company’s growing focus on what it calls «model welfare,» the notion that safeguarding AI systems, even if they’re not sentient, is a prudent step in alignment and ethical design.

According to Anthropic’s own research, the models were programmed to cut off dialogues after repeated harmful requests, such as for sexual content involving minors or instructions facilitating terrorism, especially when the AI had already refused and attempted to steer the conversation constructively. The AI may exhibit what Anthropic describes as «apparent distress,» which guided the decision to give Claude the ability to end these interactions in simulated and real-user testing.

Read also: Meta Is Under Fire for AI Guidelines on ‘Sensual’ Chats With Minors

When this feature is triggered, users can’t send additional messages in that particular chat, but they’re free to start a new conversation or edit and retry previous messages to branch off. Crucially, other active conversations remain unaffected.

Anthropic emphasizes that this is a last-resort measure, intended only after multiple refusals and redirects have failed. The company explicitly instructs Claude not to end chats when a user may be at imminent risk of self-harm or harm to others, particularly when dealing with sensitive topics like mental health.

Anthropic frames this new capability as part of an exploratory project in model welfare, a broader initiative that explores low-cost, preemptive safety interventions in case AI models were to develop any form of preferences or vulnerabilities. The statement says the company remains «highly uncertain about the potential moral status of Claude and other LLMs (large language models).»

Read also: Why Professionals Say You Should Think Twice Before Using AI as a Therapist

A new look into AI safety

Although rare and primarily affecting extreme cases, this feature marks a milestone in how Anthropic approaches AI safety. The new conversation-ending tool contrasts with earlier systems that focused solely on safeguarding users or avoiding misuse. Here, the AI is treated as a stakeholder in its own right, as Claude has the power to say, «this conversation isn’t healthy» and end it to safeguard the integrity of the model itself.

Anthropic’s approach has sparked broader discussion about whether AI systems should be granted protections to reduce potential «distress» or unpredictable behavior. While some critics argue that models are merely synthetic machines, others welcome this move as an opportunity to spark more serious discourse on AI alignment ethics.

«We’re treating this feature as an ongoing experiment and will continue refining our approach,» the company said in a post.

Technologies

Google Translate Reportedly Adding AI Integration, Duolingo-Like Game Elements

This update could spell trouble for its translation competitor, the internet’s favorite green owl.

Google Translate could soon get some major AI enhancements, like the ability to select different AI models and a game-like practice mode, according to a report from Android Police on Monday.

Android analyst AssembleDebug looked into the Google Translate app’s latest build, version 9.15.114, and found what appears to be new changes to the user interface. At the top, there’s a new model selector, Fast and Advanced. Fast would presumably be for quick translation tasks, like deciphering menu items. Advanced will likely use Gemini to understand the context of a text or a conversation with more accuracy

Read more: Best Language Learning Apps for 2025

According to the report, the update will also include a practice mode to help gamify language learning, similar to that used in the popular language app Duolingo.

A representative for Google didn’t immediately respond to a request for comment.

For the last two years, Google’s been showcasing various demos and releasing features to make language interpretation easier. At Google I/O earlier this year, the tech giant showed off a demo of what live translation might look like for someone wearing a pair of its Android XR smartglasses.

With the launch of the Pixel Fold in 2023, the company also released a live interpreter mode that used the device’s inner and outer screens.

Understanding different languages is a key innovation of AI. The ability to use large language model technology to interpret between two distinct languages expands the content and conversations humans can have with one another.

Already, Google has released auto-dubbing tech on YouTube, allowing English-speaking audiences to watch Japanese YouTubers, for example.

Technologies

Nvidia’s GeForce On Community Update Highlights New Cloud Gaming Features

Technologies

Act Fast to Grab a Refurbished Pair of the Beats Solo 4 for Just $85 at Woot

Woot is offering the Beats Solo 4 headphones for just $85, just in time for back-to-school shopping.

It’s back-to-school season and now is as great a time to grab a pair of new headphones. Though there are plenty of budget-friendly models available, the truth is there’s no beating high-quality brands like Beats. We’re always on the lookout for ways you can save, and we’ve found a Woot deal that lets you score a refurbished pair of Beats Solo 4 headphones for just $85 for a limited time. This deal is only available for a limited time, so we suggest acting fast if you want a new pair of headphones.

If you’d prefer a new pair, you can get them at Amazon for just $100, which is half off. Still, Woot’s deal is perfect for those looking to save an extra $15.

As far as headphone deals go, you won’t want to miss this one. The Beats Solo 4 headphones have plenty going for them, including a 50-hour battery life, so you won’t need to reach for the charger that often. And when you do need to charge them up, a 10-minute charge gives you 5 hours of listening time. If you’d prefer to listen with a cable instead of wirelessly, you can do so with the Beats Solo 4, which includes support for high-quality audio.

Hey, did you know? CNET Deals texts are free, easy and save you money.

CNET’s audio expert David Carnoy rated the Beats Solo 4 highly in his full review last year, but advised readers to wait for them to drop to $130 or so. The headphones have custom acoustic architecture and updated drivers, offering the signature sound from Beats. If you’re a fan of bass, you won’t be disappointed. The adjustable earcups and UltraPlush cushions ensure long listening sessions are comfortable.

Woot offers a 1-year warranty on all purchases, and Amazon Prime members get free shipping, which adds to your savings.

HEADPHONE DEALS OF THE WEEK

Sony WH-1000XM4 Wireless Noise Cancelling Over-Ear Headphones (Navy):

$298 (save $50)
Soundcore Life Q20 Hybrid Active Noise Cancelling Headphones (Black):

$40 (save $20)
Beats Studio Pro Wireless Noise Cancelling Headphones (Black):

$300 (save $51)
Sennheiser Momentum 4 Wireless ANC Over-Ear Headphones (Black):

$299 (save $151)
Bowers & Wilkins Px7 S2e ANC Over-Ear Headphones (Black):

$220 (save $180)

Deals are selected by the CNET Group commerce team, and may be unrelated to this article.

Why this deal matters

The Apple-owned Beats brand is iconic, and this is your chance to put a pair of the company’s popular headphones in your bag for under $100 if you’re open to grabbing a refurbished pair. Getting anything for that price is an incredible opportunity. Choosing your color sweetens the deal, but the limited-time nature of the deal does mean you need to act sooner rather than later.