search
HomeTechnology peripheralsIt IndustryEnhancing artificial intelligence safety: exploring safety 'braking' technology

According to news on May 26, Microsoft today released a document called "Governing AI: A Blueprint for the Future" report puts forward its own views and suggestions on the governance of generative artificial intelligence.

The report’s foreword was written by Microsoft President Brad Smith, who summarized five key takeaways aimed at strengthening the safety of artificial intelligence and making it more acceptable to the public.

First, the report recommends implementing and establishing a new government-led AI security framework. The National Institute of Standards and Technology's framework provides guidance to governments and businesses.

Enhancing artificial intelligence safety: exploring safety braking technology

In addition, the report proposes that in artificial intelligence systems, security measures to control important infrastructure should be effectively deployed. These measures will involve fail-safe systems to ensure the overall safety of the system, while emphasizing effective supervision of personnel, development of resilience and development of robustness.

The report emphasizes the establishment of a universal legal and regulatory system to meet the challenges brought by artificial intelligence technology. Smith suggests that existing laws and regulations will need to be revised to accommodate the development of artificial intelligence, and that a new government agency may need to be established to regulate the field.

To increase transparency and ensure access to artificial intelligence technology for academia and nonprofit organizations, Microsoft will publish an annual report detailing its work on artificial intelligence transparency and commit to further expanding artificial intelligence research. Collaboration, spanning academia and non-profit organizations.

Ultimately, the report highlights the need to view AI as an effective tool to address the social challenges posed by emerging technologies, based on public-private partnerships. Smith said now is the time for important work to be done by harnessing artificial intelligence to protect democracy and fundamental rights, promote inclusive growth, and harness its power to advance a sustainable planet.

As far as the editor is aware, the report released by Microsoft provides useful suggestions and guidance for governing generative artificial intelligence. This research provides strong support and puts forward a series of important perspectives and suggestions for relevant institutions and enterprises to formulate artificial intelligence safety and regulatory strategies. The report highlights the importance of transparency, cooperation and legal frameworks as factors that will have a positive impact on the future development of artificial intelligence.

The above is the detailed content of Enhancing artificial intelligence safety: exploring safety 'braking' technology. For more information, please follow other related articles on the PHP Chinese website!

Statement
This article is reproduced at:ITBear科技资讯. If there is any infringement, please contact admin@php.cn delete
Behind the first Android access to DeepSeek: Seeing the power of womenBehind the first Android access to DeepSeek: Seeing the power of womenMar 12, 2025 pm 12:27 PM

The rise of Chinese women's tech power in the field of AI: The story behind Honor's collaboration with DeepSeek women's contribution to the field of technology is becoming increasingly significant. Data from the Ministry of Science and Technology of China shows that the number of female science and technology workers is huge and shows unique social value sensitivity in the development of AI algorithms. This article will focus on Honor mobile phones and explore the strength of the female team behind it being the first to connect to the DeepSeek big model, showing how they can promote technological progress and reshape the value coordinate system of technological development. On February 8, 2024, Honor officially launched the DeepSeek-R1 full-blood version big model, becoming the first manufacturer in the Android camp to connect to DeepSeek, arousing enthusiastic response from users. Behind this success, female team members are making product decisions, technical breakthroughs and users

DeepSeek's 'amazing' profit: the theoretical profit margin is as high as 545%!DeepSeek's 'amazing' profit: the theoretical profit margin is as high as 545%!Mar 12, 2025 pm 12:21 PM

DeepSeek released a technical article on Zhihu, introducing its DeepSeek-V3/R1 inference system in detail, and disclosed key financial data for the first time, which attracted industry attention. The article shows that the system's daily cost profit margin is as high as 545%, setting a new high in global AI big model profit. DeepSeek's low-cost strategy gives it an advantage in market competition. The cost of its model training is only 1%-5% of similar products, and the cost of V3 model training is only US$5.576 million, far lower than that of its competitors. Meanwhile, R1's API pricing is only 1/7 to 1/2 of OpenAIo3-mini. These data prove the commercial feasibility of the DeepSeek technology route and also establish the efficient profitability of AI models.

Top 10 Best Free Backlink Checker Tools in 2025Top 10 Best Free Backlink Checker Tools in 2025Mar 21, 2025 am 08:28 AM

Website construction is just the first step: the importance of SEO and backlinks Building a website is just the first step to converting it into a valuable marketing asset. You need to do SEO optimization to improve the visibility of your website in search engines and attract potential customers. Backlinks are the key to improving your website rankings, and it shows Google and other search engines the authority and credibility of your website. Not all backlinks are beneficial: Identify and avoid harmful links Not all backlinks are beneficial. Harmful links can harm your ranking. Excellent free backlink checking tool monitors the source of links to your website and reminds you of harmful links. In addition, you can also analyze your competitors’ link strategies and learn from them. Free backlink checking tool: Your SEO intelligence officer

Midea launches its first DeepSeek air conditioner: AI voice interaction can achieve 400,000 commands!Midea launches its first DeepSeek air conditioner: AI voice interaction can achieve 400,000 commands!Mar 12, 2025 pm 12:18 PM

Midea will soon release its first air conditioner equipped with a DeepSeek big model - Midea fresh and clean air machine T6. The press conference is scheduled to be held at 1:30 pm on March 1. This air conditioner is equipped with an advanced air intelligent driving system, which can intelligently adjust parameters such as temperature, humidity and wind speed according to the environment. More importantly, it integrates the DeepSeek big model and supports more than 400,000 AI voice commands. Midea's move has caused heated discussions in the industry, and is particularly concerned about the significance of combining white goods and large models. Unlike the simple temperature settings of traditional air conditioners, Midea fresh and clean air machine T6 can understand more complex and vague instructions and intelligently adjust humidity according to the home environment, significantly improving the user experience.

Another national product from Baidu is connected to DeepSeek. Is it open or follow the trend?Another national product from Baidu is connected to DeepSeek. Is it open or follow the trend?Mar 12, 2025 pm 01:48 PM

DeepSeek-R1 empowers Baidu Library and Netdisk: The perfect integration of deep thinking and action has quickly integrated into many platforms in just one month. With its bold strategic layout, Baidu integrates DeepSeek as a third-party model partner and integrates it into its ecosystem, which marks a major progress in its "big model search" ecological strategy. Baidu Search and Wenxin Intelligent Intelligent Platform are the first to connect to the deep search functions of DeepSeek and Wenxin big models, providing users with a free AI search experience. At the same time, the classic slogan of "You will know when you go to Baidu", and the new version of Baidu APP also integrates the capabilities of Wenxin's big model and DeepSeek, launching "AI search" and "wide network information refinement"

Prompt Engineering for Web DevelopmentPrompt Engineering for Web DevelopmentMar 09, 2025 am 08:27 AM

AI Prompt Engineering for Code Generation: A Developer's Guide The landscape of code development is poised for a significant shift. Mastering Large Language Models (LLMs) and prompt engineering will be crucial for developers in the coming years. Th

Building a Network Vulnerability Scanner with GoBuilding a Network Vulnerability Scanner with GoApr 01, 2025 am 08:27 AM

This Go-based network vulnerability scanner efficiently identifies potential security weaknesses. It leverages Go's concurrency features for speed and includes service detection and vulnerability matching. Let's explore its capabilities and ethical

See all articles

Hot AI Tools

Undresser.AI Undress

Undresser.AI Undress

AI-powered app for creating realistic nude photos

AI Clothes Remover

AI Clothes Remover

Online AI tool for removing clothes from photos.

Undress AI Tool

Undress AI Tool

Undress images for free

Clothoff.io

Clothoff.io

AI clothes remover

AI Hentai Generator

AI Hentai Generator

Generate AI Hentai for free.

Hot Article

Hot Tools

PhpStorm Mac version

PhpStorm Mac version

The latest (2018.2.1) professional PHP integrated development tool

MantisBT

MantisBT

Mantis is an easy-to-deploy web-based defect tracking tool designed to aid in product defect tracking. It requires PHP, MySQL and a web server. Check out our demo and hosting services.

SublimeText3 Linux new version

SublimeText3 Linux new version

SublimeText3 Linux latest version

SecLists

SecLists

SecLists is the ultimate security tester's companion. It is a collection of various types of lists that are frequently used during security assessments, all in one place. SecLists helps make security testing more efficient and productive by conveniently providing all the lists a security tester might need. List types include usernames, passwords, URLs, fuzzing payloads, sensitive data patterns, web shells, and more. The tester can simply pull this repository onto a new test machine and he will have access to every type of list he needs.

EditPlus Chinese cracked version

EditPlus Chinese cracked version

Small size, syntax highlighting, does not support code prompt function