Recently, several major American artificial intelligence companies made a solemn pledge to the White House, stating that they will prevent their AI products from being used to generate non-consensual deepfake pornography and child sexual abuse materials. Six companies, including Adobe, Anthropic, Cohere, Microsoft, OpenAI, and the open-source web data repository Common Crawl, signed this non-binding commitment to ensure that their products are not misused to create harmful images of gender violence.

Cosmos Institute has announced a program aimed at cultivating a new generation of tech talent and entrepreneurs, focusing on fostering individuals who can think deeply and uphold humanistic values in the age of artificial intelligence. The founding members of the institute include Jack Clark of Anthropic and former Department of Defense technology expert Brandon McCall. The core goal is to create AI technologies that respect human dignity and mitigate the negative impacts of automation. The institute has established the 'Human-Centered AI Lab' at the University of Oxford, led by philosopher Philip Kouroupos, with the aim of integrating philosophical principles.

Anthropic has released Claude Enterprise, an enterprise-grade AI assistant designed to enhance team efficiency. By integrating GitHub, expanding the context window, and providing enterprise-level security features, it helps companies better leverage their internal knowledge assets. This product can handle large volumes of documents and code, accelerating code repository synchronization, feature iteration, and issue debugging, while ensuring data security. Market and product managers can use it to generate marketing strategies and interactive prototypes, improving cross-department communication efficiency. Early user feedback has been positive.

Anthropic has announced that it will release the Artifacts system prompts for the Claude series models within a few weeks, responding to researchers' concerns about transparency. The Artifacts feature was recently opened to all users, allowing code snippets to run alongside the chat interface. The release of the system prompts aims to enhance user understanding of model behavior and promote transparency in AI development. The Claude 3.5 Sonnet version emphasizes accurate and concise answers, while Claude 3 Opus has been updated.

Amazon has announced the launch of an improved version of its voice assistant, named 'Remarkable Alexa', which is primarily based on Anthropic's Claude AI model. This move aims to enhance Alexa's performance and address accuracy issues in understanding user commands and needs. The new assistant will introduce several practical features, such as daily AI-generated news summaries, a child-specific chatbot, and conversational shopping tools. Expected to be released in mid-October, it will operate on a subscription model with a monthly fee of approximately $5 to $10.

The AI chatbot Claude from Anthropic has recently come under scrutiny due to user feedback suggesting a decline in performance, including reduced memory and deteriorating coding abilities. A post on Reddit claimed that Claude's intelligence level seems to have decreased. In response, Anthropic executive Alex Albert stated that the company's investigation found no widespread issues and confirmed that no changes had been made to Claude's model or reasoning pipeline. To increase transparency, the company has released the system prompts for the Claude model.

Anthropic has recently made public the system prompts for the Claude model, including the Claude 3.5 Opus, Sonnet, and Haiku versions. This is a significant breakthrough for the AI industry, as most vendors typically keep this information confidential. This move enhances AI transparency, as the behavior rules and personality traits of Claude are detailed through these prompts. Generative AI models generate text based on statistical predictions, and system prompts are used to guide model behavior to avoid inappropriate actions. Anthropic's aim with this initiative is to

Claude, the AI large model developed by Anthropic, has introduced a LaTeX formula rendering feature that significantly enhances the clarity and professionalism of mathematical expressions. Key highlights include improved consistency in formatting, standardized display of mathematical equations and expressions, making them easier to understand. The incorporation of LaTeX greatly enhances the clarity of answers, especially suitable for scenarios requiring solutions to mathematical problems. This feature meets the high demands of academic and professional fields, allowing users to quickly generate complex mathematical expressions with simple command inputs.

A San Francisco-based AI startup, Anthropic, is facing a lawsuit from a group of writers for allegedly using a large number of pirated books in the training of its chatbot Claude. This marks the first time writers have filed a lawsuit against Anthropic, accusing its actions of 'mocking its lofty goals' and utilizing pirated resources to build AI products. The complaint states that Anthropic is attempting to profit from the human expression and creativity behind each work. The plaintiffs include writers Andrea Bartz and Charles Graeb.

The Anthropic API introduces a new prompt caching feature aimed at optimizing the user experience with the Claude model. This feature allows users to cache frequently used contextual information between API calls, significantly reducing the costs and latency of long prompts by up to 90% and 85%, respectively. Currently, it supports the public beta versions of Claude 3.5 Sonnet and Claude 3 Haiku, with plans to expand to Claude 3 Opus in the future. It is suitable for use cases such as conversational agents, coding assistants, handling large documents, and agent searches.

Anthropic has announced an expansion of its vulnerability reward program aimed at testing a 'next-generation AI safety mitigation system,' primarily focusing on identifying and defending against 'universal jailbreak attacks.' Special attention is given to high-risk areas, including CBRN defense and cybersecurity. Participants will have the opportunity to engage with the new safety system ahead of time, identifying vulnerabilities or bypassing security measures, with rewards of up to $15,000. This initiative aims to enhance the safety of AI systems by attracting security researchers to collaboratively discover and fix potential threats, setting a benchmark for safety in the AI industry.

The UK's Competition and Markets Authority (CMA) is investigating Amazon's $4 billion investment in the AI startup Anthropic, concerned that the partnership may affect market competition in the UK. The investigation is based on information gathered by the CMA, which believes the collaboration could have adverse effects. Amazon has expressed dissatisfaction, stating that the partnership meets market standards. The CMA has collected public opinions and plans to assess the impact of the partnership on the market in the coming weeks, with a final decision to be announced on October 4. The investigation involves Anthropic's use of Amazon Web Services.

OpenAI co-founder John Schulman has left the company to join the AI startup Anthropic. This move reflects the growing trend of talent mobility and intensifying competition in the industry. With extensive experience in the field of artificial intelligence, Schulman's addition to Anthropic may drive innovation and attract industry attention. As an emerging force, Anthropic stands to benefit from Schulman's fresh perspective and energy. This decision is seen as a microcosm of the increasing talent movement and competition within the AI sector, indicating a future that will be even more dynamic and competitive.

For the first time in its annual 10-K report, Microsoft listed OpenAI as a competitor, raising industry concerns about the potential changes in their partnership. Microsoft has a long-standing collaboration with OpenAI, investing $13 billion and providing exclusive cloud services. However, OpenAI's SearchGPT feature positions it as a potential competitor to Microsoft in the search domain. An ongoing U.S. antitrust investigation may impact their relationship. Microsoft's decision to relinquish its observer seat on OpenAI's board suggests possible changes in their partnership. However, experts point out that partners and competitors often coexist in the tech industry.

The UK Competition and Markets Authority (CMA) has announced a preliminary investigation into the relationship between Google and the AI startup Anthropic. This investigation is currently in its early stages and aims to assess whether the partnership could potentially weaken market competition and affect the UK market. Anthropic focuses on developing safe, transparent, and risk-controlled AI systems, having raised nearly $10 billion, with Google as one of its main investors, contributing over $2.3 billion. The CMA is inviting all 'interested parties' to participate in the consultation until August 2024.

The EU has initiated a consultation on rules for providers of General AI Models (GPAI), aiming to provide guidance for compliance with the EU AI Act legal obligations and ensure the 'trustworthiness' of GPAI. The consultation is divided into three parts, focusing on transparency, risk classification and assessment, and the review and monitoring of codes of conduct, to establish relevant behavioral guidelines. Participants include GPAI providers, businesses, civil society, rights holders, and academic experts, whose feedback will influence the formulation of the behavioral guidelines, particularly regarding the summary of model training content.

Recently, an investigation revealed that hundreds of websites attempted to block the AI company Anthropic from scraping their content, but due to outdated directives, they ended up blocking the wrong bots. This phenomenon highlights the challenges website owners face in coping with the ever-evolving ecosystem of AI crawlers.

OpenAI and Anthropic may face losses of up to billions of dollars this year, primarily due to the high costs of training and operating AI models. OpenAI's AI training costs are expected to reach $3 billion, with personnel costs potentially hitting $1.5 billion and rental of Microsoft servers costing nearly $4 billion. Anthropic's situation is more severe, with projected expenses exceeding $2.7 billion, while its revenue is only one-fifth to one-tenth of OpenAI's. The costs of developing and operating AI models are facing intense competition, especially in multimodal scenarios.