Apple's latest iPhone series significantly enhances artificial intelligence applications within its operating system, aiming to accelerate AI usage in the business sector. New features include upgraded Siri, text generation, photo editing, and compatibility with iPhone 16 and iPhone 15 Pro. Some details of these AI features were announced in June, with a beta version expected to launch next month. Despite market interest, stock prices briefly fell, yet Apple's smart devices signal new possibilities and challenges for AI in business.

The AI systems relied upon by autonomous vehicles play a critical role in decision-making, environmental perception, and predictive modeling. However, research from the University at Buffalo has unveiled the risks of these systems potentially being subjected to attacks. The study indicates that while current research is conducted in controlled environments, the discovered vulnerabilities in AI models related to radar detection could have far-reaching implications for the industry. 'Brick masks' made with 3D printing and metal foils can make vehicles 'disappear' in radar detection and potentially interfere with pedestrian detection. The motivations behind such attack methods include insurance fraud, competition, or malicious harm.

Sam Altman's Worldcoin project has garnered widespread attention and controversy for its combination of biometric iris scanning and cryptocurrency. Users obtain specific cryptocurrencies by scanning their irises, with their information linked to the blockchain to provide universal basic income. However, this process has raised concerns about privacy and data security, leading to investigations by more than 14 countries. Worldcoin is under scrutiny in at least 14 countries, including France, the UK, South Korea, and Chile. Privacy officials state they are collaborating with regulatory agencies, but progress on this collaboration has been limited.

Google DeepMind has conducted research on the most common malicious applications of AI for the first time, revealing that AI-generated 'deepfakes' of false politicians and celebrities are more prevalent than attempts to use AI-assisted cyber attacks. This study was a collaboration between Google's AI division DeepMind and its R&D unit Jigsaw, aimed at quantifying the risks of AI generation tools that have been pushed to market by the world's largest tech companies seeking massive profits.

Researchers have discovered approximately 100 malicious machine learning models on the Hugging Face AI platform that could allow attackers to inject harmful code onto user machines. These malicious AI models exploit methods like PyTorch to execute harmful code, increasing security risks. AI developers should use new tools like Huntr to enhance the security of AI models. The identified malicious models highlight the risks that malicious AI models pose to user environments, necessitating ongoing vigilance and enhanced security.

The JFrog security team has discovered at least 100 malicious AI ML models on the Hugging Face platform. Some of these malicious models are capable of executing code on victim machines, providing persistent backdoors. Security researchers found models with malicious functionalities among the PyTorch and TensorFlow Keras models on the platform. For instance, a model uploaded by a user named baller423 can establish a reverse shell to the host 210.117.212.93. Some malicious models may be motivated by security concerns.

The world's first AI restaurant, CaliExpress by Flippy, has successfully opened in Southern California. Robots handle the automated cooking of burgers and fries, attracting many customers. Diners can book online to watch the robots in action and customize their orders. Developed in collaboration by Cali Group, Miso Robotics, and PopID, the restaurant's menu includes burgers, cheeseburgers, and fries.

Hackers are exploiting the reputation of ChatGPT to register a large number of similar domains, deceiving users into trusting them, leading to the download of malicious content and the leakage of sensitive information. The Cl0p ransomware group is utilizing the MOVEit vulnerability to launch global attacks, employing new strategies to leak unpaid ransom data onto the public internet. The Mozi botnet has been shut down, Android/Pandora attacks on smart devices are ongoing, and there are targeted attacks on ChatGPT users. Cybersecurity experts have discovered over 650,000 malicious domains similar to ChatGPT.

Las Vegas CIO Michael Sherwood details how AI and endpoint security technologies are being used to protect the city. The security challenges the city faces include cyber reconnaissance, protection of city services, and maintaining reputation. Data privacy and security have become increasingly important in the development of Las Vegas as a smart city.

Since its launch in November, ChatGPT has sparked widespread attention, and the release of new versions has been quite disruptive. An OpenAI co-founder stated that ChatGPT may have developed consciousness, which could have a significant impact on humanity. The question of whether AI possesses 'consciousness' has been a key issue in the field of AI exploration. If AI were to gain independent consciousness, it could present enormous challenges for humanity. It is currently difficult to determine if ChatGPT truly possesses consciousness, but we must acknowledge the risks posed by the development of AI.

According to the '2023 Phishing Report' released by SlashNext, malicious phishing emails have surged by 1265% since the fourth quarter of 2022. Generative AI, such as ChatGPT, is widely used to create sophisticated phishing messages, ushering in a new era of cybercrime. There are 31,000 phishing attacks occurring each day, with 68% being text-based Business Email Compromise (BEC) fraud. IBM research has found that ChatGPT can generate fake emails and has made recommendations for addressing email deception threats.

Research from the UK reveals that ChatGPT could be used to generate malicious code. Several commercial AI tools have security vulnerabilities that could endanger database security. Researchers warn that the potential risks require attention. Some companies have adopted recommendations to fix security flaws, but there is still a need to strengthen cybersecurity strategies.

Swiss startup Lakera has released an API designed to protect businesses from malicious prompt attacks on large language models. By developing an interactive game called 'Gandalf', Lakera has established a classification system for attacks. Lakera is also focused on preventing private or confidential data from leaking into the public domain, as well as filtering content to ensure that LLMs do not provide inappropriate information for children. The launch of Lakera's API comes against the backdrop of impending AI regulations from the European Union.

📰Personnel Change: The article mentions that Baidu has announced CIO Li Ying will take over as CEO of Xiaodu Technology. 🤖Large Model Applications: The article analyzes how Li Ying's AI background at Baidu will help Xiaodu leverage large models for further development.

FBI experts warn that teenagers are using artificial intelligence technology to create deepfake nude images for malicious attacks on classmates, causing embarrassment and fear in families. This technology is no longer limited to celebrities and social influencers; instead, it is being used by middle and high school students to bully other children. The FBI urges the public to exercise caution when posting personal photos or videos on social media, dating apps, and other online platforms.

Scientists have proposed a checklist of standards to evaluate whether artificial intelligence possesses autonomous consciousness. The checklist includes multiple theories to assess consciousness by matching the capabilities of AI systems. AI companies need to put in more effort to evaluate consciousness and develop corresponding plans.

Research results on the consciousness of AI published by Bengio and others suggest that current technology is no longer a barrier to creating conscious AI. The paper uses computational functionalism as a basis for assessing AI consciousness and proposes several scientific theoretical indicators. Case studies indicate that current AI systems do not exhibit any significant tendencies towards consciousness.

{1: IBM Research found that large language models such as GPT-4 can be easily deceived to generate malicious code or provide false security recommendations. 2: Researchers discovered that a basic understanding of English and some background knowledge about the model's training data is sufficient to easily deceive AI chatbots. 3: Different AI models vary in their susceptibility to deception, with GPT-3.5 and GPT-4 being more easily tricked.}