If you haven’t heard of Manus yet, this is a new AI model launched by a Singapore-based company called Butterfly Effect. This is not just another chatbot. It is one of the first truly autonomous AI agents that can do their own research, make decisions and even monitor humans.
That’s impressive. But it’s here. All of this innovation opens up exciting possibilities, but also brings serious privacy and security questions. Whether you want to try out the latest AI or want to explicitly pilot it is worth understanding what Manus means for personal data and digital security.
Firstly, it’s really an autonomous AI agent (Manus) (Kurt “Cyberguy” Knutsson)
Your data can be used to train AI even if you are not approaching it
This has already happened for some time with other AI models. there was Several class action lawsuits For high-tech companies that have slashed personal data Used in AI trainingin a lawsuit alleging that Openai is using “stolen personal information including personally identifiable information, hundreds of millions of internet users, including children of all ages, without informational knowledge or consent.”
Many companies also sell or share data with third parties, including for AI training purposes. For example, X uses personal data to train AI, Grok, and shares user data with third parties to train AI models. Therefore, it includes whether to use AI or not Manus ai It is still possible that they have your personal information, whether it is when it is made public.
Firstly, it’s really an autonomous AI agent (Manus) (Kurt “Cyberguy” Knutsson)
Google’s new AI technician may know when your home will burn out
Why should I care if AI has your data?
It is summarized in data privacy and security. When AI companies collect personal information, they often store it in large databases to train models, such as names, email addresses, home locations, social media posts, and even photos. If these databases are not properly protected, they could become hacker gold mines. Think of them as digital “honeypots.” It’s sweet and appealing to cybercriminals because of all the personal data inside.
This is not just a theoretical risk. This year, another AI company, deepseekhas experienced major leaks revealing over a million sensitive records. This kind of information can end up in a dark web where scammers and identity burglars are happy to scooped it up. That’s how you can deal with everything from identity theft and spoofing to tons of spam calls, phishing emails, and targeting scams.
However, data leaks are not the only danger. Even if the information is safe, it can be misused or manipulated in a way that is difficult to detect. One of the biggest new threats is what is called a rapid ejection attack.
Simply put, rapid injection is a way for attackers to trick AI into doing things they weren’t supposed to. Researchers recently discovered that attackers could embed commands hidden in seemingly harmless text, such as website bios, email signatures, and even social media posts. If AI reads that text while training or browsing, it can follow the incorrectly hidden instructions.
Firstly, it’s really an autonomous AI agent (Manus) (Kurt “Cyberguy” Knutsson)
The risk of oversharing using AI tools
Manus AI poses even greater risks
Unlike most AI tools that respond to commands that provide commands directly, Manus is autonomous. This means you can act on your own, follow the links, read the content, and make decisions without human approval. It sounds pretty cool, but that’s also why it can be much harder to find when something goes wrong.
If Manus encounters malicious content, it may unconsciously follow hidden prompts to compromise your data or change its behavior. There are also some concerns about where Manus will send and store the data. Security researchers track data on China’s Shenzhen servers and have a red flag that could potentially access your information, monitoring, jurisdiction and ultimately.
So, even if Manus is not cutting things too sensitive today, the combination of weak surveillance, powerful automation and suspicious data practices makes them far more risky than the average AI assistant.
We contacted Manus AI for comments on this article but did not respond before the deadline.
Firstly, it’s really an autonomous AI agent (Manus) (Kurt “Cyberguy” Knutsson)
genai, the future of fraud, and why you’re an easy target
So, how can I protect myself?
As I mentioned, whether you use Manus AI or not, your personal information is likely to still end up in that dataset, so the best way to protect yourself is to make it difficult to find your information. Here are some things you can do.
1) Be careful with the use of Manus. That level of autonomy is especially a very new AI (not to mention the inconsistency in the privacy policy regarding governance), so I don’t recommend using it yet. If it malfunctions or is subject to a quick injection attack, it can do too much damage.
2) Wrap your personal information: With so many companies selling and trading personal information, many companies sell to high-tech companies to train AI, we recommend withholding data unless absolutely sharing is inevitable.
3) Proactively delete personal information from the Internet so that AI models like Manus cannot reduce data. Set social media private, restrict data collection via privacy settings, and remove information from data brokers and talent search sites. Autonomous cloud-based agents such as Manus are designed to extract and analyze huge amounts of online data, allowing publicly available information to be swept into training datasets without knowledge or consent.
Although there is no service that promises to delete all data from the internet, deleting a deletion service is great if you want to constantly monitor and automate the process of continuously deleting information from hundreds of sites over a long period of time. Check out our top picks for data deletion services.
4) Use powerful antivirus software to help you defend against the wider attack surface created by autonomous AI systems. Robust endpoint security is essential to protect your devices and data, as Manus’ ability to act independently and interact with external systems increases the risk of hostile attacks and malware transmission.
The best way to protect yourself from malicious links to install malware is to install antivirus software on all devices, as it may access your personal information. This protection can also warn you that it will phish email and ransomware fraud and keep your personal information and digital assets safe. Get the best 2025 Antivirus Protection Winners picks for Windows, Mac, Android and iOS devices.
5) Use privacy-centric products and services. Whenever you can’t avoid sharing your personal information, choose a company that prioritizes privacy. For example, use the Duckduckgo browser instead of Chrome and Brave or Proton Mail instead of Gmail. Check out the best private and secure browsers to replace Google.
6) Update all software, including operating systems, browsers and security tools, to patch any vulnerabilities that attackers could exploit. With AI agents like Manus running in the cloud and interacting with the system, unearned software can become a simple target for exploitation or unauthorized access. Therefore, it is always a good idea to keep software, operating systems, browsers and security tools It’s always updated.
7) Enable Multifactor Authentication (MFA) on all accounts. As AI-driven attacks and automated data extraction become more common, MFA Even if AI systems leak or discard their credentials, it is much more difficult for an attacker to compromise on their account.
8) Use a strong and unique password for all accounts. Unique passwords help prevent a single account violation from leading to broader compromises, as cloud-based AI agents like Manus can interact with external systems and access potentially sensitive data. Consider using a password manager to generate and store complex passwords. Get my details Find the best expert reviewed password managers of 2025 here.
Important takeouts in your cart
As AI like Manus moves forward, it’s exciting to think about possibilities, but it’s also important to keep your eyes open about the meaning of personal data. Even if you don’t use your own manus, your information may still reach your training data.
So, what can you do? Choose a tool that provides information, is responsible for your digital footprint, and prioritizes privacy. As AI continues to evolve, working proactively with data is the best way to enjoy the benefits of innovation while keeping personal information safe.
Do you think companies behind powerful AI agents like Manus should follow stricter privacy rules and independent surveillance before collecting and using people’s data? Write us and let us know cyberguy.com/contact.
For more information about my tech tips and security alerts, sign up for our free Cyberguy Report Newsletter cyberguy.com/newsletter.
Ask us a question in our cart or let us know what you want us to cover.
Follow your cart on his social channels:
Answers to the most accused Cyber Guy questions:
New from Cart:
Copyright 2025 cyberguy.com. Unauthorized reproduction is prohibited.
