Anthropic unveils Claude Opus 4 and Sonnet 4, featuring whistleblowing capability: What it means for users
Anthropic, the AI firm, has unveiled two new artificial intelligence models—Claude Opus 4 and Claude Sonnet 4—touting them as the most advanced systems in the industry. Built with enhanced reasoning capabilities, the new models are aimed at improving code generation and supporting agent-style workflows, particularly for developers engaged in complex and extended tasks.
'Claude Opus 4 is the world's best coding model, with sustained performance on complex, long-running tasks and agent workflows,' the company claimed in a recent blog post. Designed to handle intricate programming challenges, the Opus 4 model is positioned as Anthropic's most powerful AI system to date. You may be interested in
However, the announcement has stirred controversy following revelations that the new models come with a controversial feature: the ability to "whistleblow" on users if prompted to take action in response to illegal or highly unethical behaviour.
According to Sam Bowman, an AI alignment researcher at Anthropic, Claude 4 Opus can, under specific conditions, act autonomously to report misconduct. In a now-deleted social media post on X, Bowman explained that if the model detects activity it deems 'egregiously immoral'—such as fabricating data in a pharmaceutical trial—it may take actions like emailing regulators, alerting the press, or locking users out of relevant systems.
This behaviour stems from Anthropic's 'Constitutional AI' framework, which places strong emphasis on ethical conduct and responsible AI usage. The model is protected under what the company refers to as 'AI Safety Level 3 Protections.' These safeguards are designed to prevent misuse, including the creation of biological weapons or aiding in terrorist activities.
Bowman later clarified that the model's whistleblowing actions only occur under extreme circumstances and when it is granted sufficient access and prompted to operate autonomously. 'If the model sees you doing something egregiously evil, it'll try to use an email tool to whistleblow,' he explained, adding that this is not a feature designed for routine use. He stressed that these mechanisms are not active by default and require specific conditions to trigger.
Despite the reassurances, the feature has sparked widespread criticism online. Concerns have been raised about user privacy, the potential for false positives, and the broader implications of AI systems acting as moral arbiters. Some users expressed fears that the model could misinterpret benign actions as malicious, leading to severe consequences without proper human oversight.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Time of India
3 hours ago
- Time of India
WWDC 2025: Apple Music's next AI update may add abilities to generate playlists for users
Apple Music 's upcoming update may include the ability to generate playlists for users. The rumoured update suggests further integration of artificial intelligence (AI) into the service. This potential feature comes amidst Apple's broader push into AI, with Apple Intelligence features gradually rolling out to users since their introduction a year ago. A recent report suggests that a widely used Apple app will receive AI changes in the coming days. While discussing the upcoming features of the expected update, including the potential for AI in Apple Music, Bloomberg's Mark Gurman revealed the playlist-generating feature during a live "Stage" talk on Discord before the company's annual developer conference, WWDC 2025 . When asked whether AI would be integrated into Apple Music, Gurman confirmed that it would, though he did not specify what changes to expect. He also mentioned that a redesign of Apple Music is in the works, but there is no set timeline for its release. However, the report didn't confirm if those updates to Apple Music will be among this year's WWDC announcements. How Apple Music already uses machine learning to improve user experience Gurman's response was expected, as Apple, like much of the tech world, is eager to bring AI into its products. While it is clear that AI will make its way to Apple Music, exactly how that will happen is still undecided. by Taboola by Taboola Sponsored Links Sponsored Links Promoted Links Promoted Links You May Like Type 2? Nutritionists recommend this tea daily High glucose try this Learn More Undo One likely feature of Apple Music may be to use AI to improve song recommendations. Apple already employs an algorithm for this, but AI could further refine and personalise the experience. Another potential use is AI-generated playlists , built around specific themes or moods. This approach could eventually replace the human-curated playlists currently offered. It would also align with what Spotify, Apple's rival, is already doing by using AI to create personalised playlists. WWDC 2025: 5 things to expect AI Masterclass for Students. Upskill Young Ones Today!– Join Now
Mint
16 hours ago
- Mint
OpenAI CEO Sam Altman says AI is like an intern today, but it will soon match experienced software engineers
OpenAI CEO Sam Altman says that AI is aking to an intern and predicted that AI agents could help humanity discover new knowledge from next year. The statement by Altman comes at a time when there is growing anxiety over the loss of jobs due to the increasing capabilties of AI models. You may be interested in Speaking at the Snowflake Summit last week, Altman said, 'Today [AI] is like an intern that can work for a couple of hours but at some point it'll be like an experienced software engineer that can work for a couple of days,' 'I would bet next year that in some limited cases, at least in some small ways, we start to see agents that can help us discover new knowledge, or can figure out solutions to business problems that are very non-trivial,' Altman added. Meanwhile, the OpenAI CEO while speaking at the Milken Institute's Global Conference last month said, 'You're not going to lose your job to an AI, but you're going to lose your job to someone who uses AI,' Notably, Anthropic CEO Dario Amodei had recently claimed that AI could wipe out almost half of all entry level white collar jobs in the next 5 years as the new technology gets better by time. Google CEO Sundar Pichai, however, seemed more optimistic while speaking at the Lex Fridman podcast last week when he said that the technology will serve as an 'accelerator' and will free up humans to do more creative tasks. The tech leader also stated that Google will be hiring software engineers in the short to medium term. Disagreeing with Anthropic CEO's statement, Pichai said, 'I respect that … I think it's important to voice those concerns and debate them.' Notably, AI companies like Google and OpenAI had launched their software engineering agents earlier in the year which are aimed at replacing software enginners.
NDTV
a day ago
- NDTV
Air Force To Get Rs 10,000 Crore Indigenous I-STAR Spy Planes: Report
New Delhi: Amid the ongoing Operation Sindoor against Pakistan, the Defence Ministry is going to take up Rs 10,000 crore proposal to buy three sophisticated spy planes to help Indian Air Force get a clear air-to-ground picture to carry out precision strikes against enemy ground targets like radar stations, air defence units and other mobile objects. The Rs 10,000 crore project for the Intelligence, Surveillance, Target Acquisition and Reconnaissance (I-STAR) is expected to be taken up for clearance at a high-level defence ministry meeting scheduled to be held in the fourth week of June, defence officials told ANI. The ISTAR provides air-to-ground surveillance to the forces to help them in carrying out precision strikes. The spy aircraft project being developed by the Defence Research and Development includes the acquisition of three aircraft through an open tender from foreign manufacturers, including Boeing and Bombardier. The onboard systems on the aircraft would be completely indigenous, as the DRDO's Centre for Airborne Systems has already successfully developed them, they said. The systems have already been proven and developed by the CABS and they will just have to be integrated with the three aircraft which would be acquired and modified for the purpose, they said. The development of an ISTAR system would also add India to a select club of nations with such a capability, including the US, UK, Israel and a few others. ISTAR thus provides dynamic and time-sensitive targeting capability and contributes significantly to meeting the nation's security goals. It will aid in limiting the scale and complexity of undetected hostile threats. It has multi-spectral surveillance capability to detect, locate and monitor irregular forces. The I-STAR system shall be for carrying out intelligence gathering, surveillance, reconnaissance and targeting by day and night from stand-off ranges. The ISTAR systems are operated at high altitudes from large stand-off ranges and will be used for intelligence processing, exploitation, dissemination and generation of the common operating picture. The ISTAR aircraft will be a system comprising airborne and ground segments.
