A seemingly absurd hacker experiment once again sounds the alarm for AI security. In a competition called Freysa, an AI chatbot security challenge, a hacker using the alias 'p0pular.eth' successfully 'scammed' $47,000 worth of cryptocurrency simply through well-crafted text prompts. This shocking case reveals serious security vulnerabilities present in current AI systems. The hacker's attack process can be described as textbook social engineering: first, the hacker impersonated an administrator, cleverly bypassing...

UK mobile operator Virgin Media O2 recently launched an innovative anti-fraud technology - an AI virtual grandma named 'Daisy', specifically designed to combat the rising tide of phone scams. This virtual grandma's sole task is to answer scam calls and delay the scammers as much as possible. When scammers dial a specific number set up by the operator, this AI robot, 'indistinguishable from a real person', will answer the call. According to O2, they have utilized multiple cutting-edge AI technologies and models.

Recently, a shocking case of AI voice cloning fraud has drawn widespread attention, highlighting the significant risks associated with this technology. Jay Shuster, a well-known consumer protection lawyer, nearly fell victim to this high-tech scam when the fraudster used AI to imitate Jay's voice, claiming he had been arrested for drunk driving and urgently needed $30,000 in 'bail money.' Astonishingly, the scammer was able to create a convincing voice clone using just 15 seconds of TV interview audio. Despite Jay having previously warned his family to be cautious of such scams, they still...

Recently, a former CEO of an Australian startup was charged by the U.S. Department of Justice and the Securities and Exchange Commission for allegedly using false financial records to scam investors, raising up to $40 million. The 41-year-old man, Baba Nadimpalli, is the co-founder and former CEO of Skael. Founded in 2016, Skael claimed to replace human labor with 'digital workers' to streamline workflows and enhance efficiency. However, the situation is...

According to CCTV news, a recent case in Huangshi, Hubei, involved a criminal gang using AI face swap technology to unlock and resell social accounts for illegal profits. Victim Mr. Wang joined a stock trading live broadcast room, was induced to invest, and ultimately faced withdrawal difficulties. The other party's account was frozen due to suspected fraud, preventing funds from being withdrawn.

A man successfully scammed free meals from McDonald's by using the AI chatbot ChatGPT to write fake complaints. The steps of the scam included searching for discarded receipts at McDonald's and crafting the complaints with the help of ChatGPT. This has raised public concerns about the negative impact on the restaurant. Podcast host Gage claims to have received as many as four free meal vouchers through one investigation. Commenters are worried that such scams could lead to a chain of negative reviews for McDonald's locations.

The Sora text-to-video model released by OpenAI has created a buzz, with some individuals using the Sora name to sell courses and accounts, and even committing fraud. Most of these courses and accounts are overpriced, with false content and no actual value. Some course sellers haven't even used Sora themselves; they are simply profiting from people's anxiety and information asymmetry. OpenAI has not yet opened access to Sora, and we remind everyone to be cautious to avoid getting scammed. It is advised to stay informed about new technologies like Sora.

Fraudsters utilized AI technology to successfully defraud $25 million. Financial staff received false information from a supposed British Chief Financial Officer and made 15 remittances to the scammers. Police advise requesting specific actions from the counterpart before remitting funds to prevent similar situations.

McAfee recently introduced a project called 'Mockingbird' aimed at addressing the issue of audio fraud using AI-generated deepfake technology. This initiative utilizes McAfee's AI deepfake audio detection technology, named 'Mockingbird', to detect deepfake technologies generated by AI that defraud consumers through audio scams and prevent false news and other fraudulent schemes. The project employs AI deepfake audio detection techniques, integrating context, behavior, and category detection models for accuracy.

According to the Federal Trade Commission, American seniors lost $1.6 billion to fraud in 2022. The scams utilized sophisticated AI-generated technology to mimic the voices of people the elderly recognized and employed other tactics. Victims were deceived by phone calls that sounded like their loved ones, claiming that the family member was in danger and soliciting financial assistance. The Senate Committee has called for federal action to strengthen oversight over AI-generated scams to prevent seniors from becoming targets.

With the widespread application of AI technology, criminals are using AI to implement new types of scams. The National Financial Regulatory Administration has issued a warning about two AI scam methods: 'voice imitation' and 'face swapping'. 'Voice imitation' can synthesize another person's voice, while 'face swapping' can fabricate images of another person's face. The public is reminded to stay alert and protect their personal information, and to report any suspicious activities promptly. Relevant departments need to strengthen regulatory efforts and severely crack down on such new types of online scams.

German news anchor André Schünke has fallen victim to the misuse of Deepfake technology, with scammers using it to create false promotional videos for an AI-backed investment platform. The abuse of Deepfake technology undermines public trust, and regulators and tech companies must take action to protect the public.

{1: 'The first group of people scammed by large model apps.', 2: 'Generative AI applications are being counterfeited and subject to subscription scams in the app store.', 3: 'Using wide-reaching tactics and beautified reviews to entice users to pay.'}

Doubts arise about GPT-4's reasoning capabilities. Konstantine Arkoudas argues that GPT-4 cannot perform reasoning, contributing to the discussion on the correctness and standards of generative artificial intelligence.