First autonomous AI agent is here, but is it worth the risks?

If you haven’t heard the tingling of the manus, it is a new AI model discovered by a company based in Singapore called Butterfly Effect. This is not just another chatbot. It is one of the first truly autonomous agents of AI, which is able to do their own research, make decisions and even make plans, all with barely any human control.

Sounds impressive, doesn’t it? But there is a matter: while all this innovation opens exciting opportunities, it also brings serious questions of privacy and security. Whether you are eager to try the latest AI or you would rather be clear, it is worth understanding what Manus might mean for your personal information and digital security.

Stay protected and informed! Get safety warnings and professional technological tips – sign up for Kurt’s ‘Cyberguy Report’ now

Your data could be used to train AI, even if you are nowhere to go nearby

This has been going on with other AI models for some time. It was Several lawsuits in class against technological companies to scrape personal data for Use in AI trainingWith one lawsuit that claims to use Openi “stolen private information, including personal information, of hundreds of millions of internet users, including children of all ages, without their informed knowledge or consent.”

Many companies also sell or share data with third parties, including for AI training purposes. For example, X used personal information for training AI, GROK, and also shares user data with third parties to train the AI ​​model. So if you use AI (including Manus you have When released to public) or not, there is still a good chance of having your personal information.

Google’s new AI technology can know when your house will burn

Why would you care if there are Ai your information?

This comes down to the privacy and safety of data. When AI companies collect personal information – things like your name, e -zech addresses, home locations, social media posts, and even photos – often store them in large databases to train their models. If these databases are not properly secured, they can become gold mines for hackers. Consider them digital “honey pots”: sweet and tempting to cyber -criminal because of all personal data from the inside.

This is not just theoretical risk. Only this year, another Ai company, Deepseekexperienced a big leak that exhibited more than a million sensitive records. When this type of information is drawn, it can end up on a dark web where fraudsters and identity thieves are more than happy to take them out. Thus, people in the end deal with everything, from the theft of identity and presentation to tons of unwanted calls, Phishing E -Mail and Targeted Frauds.

But data leaks are not the only danger. Even when the information is safe, they can be abused or manipulated in ways that are difficult to detect. One of the biggest new threats is something called a quick injection attack.

Simply put, fast injection is a way to do the assault on the fraud and do something they did not need. Researchers have recently discovered that attackers can install hidden commands in a seemingly harmless text, such as a biography of the website, signatures of E -Kašte, or even fasting on social media. When the AI ​​reads that text during training or browsing, it may accidentally follow the hidden instructions, which could tell him to leak information, change the way he behaves or gives unauthorized access to private information.

The risk of over -sharing with AI tools

Manus AI represents an even greater risk

Unlike most AI tools that respond to the commands you give them directly, the manus is autonomous, which means that it can act alone, follow connections, read content and make decisions without human approval. Although that sounds pretty cool, this is also what makes it much more difficult to notice when something goes wrong.

If Manus encounters malicious content, it might unknowingly follow the hidden inquiries, endanger your information, or change his behavior, and there may be no obvious sign that anything happened until it’s too late. There are also some concerns where Manus sends and stores data. Security researchers have found data to a server in Shenzhen in China, raising red flags about supervision, jurisdiction, and which could ultimately have access to your data.

So, even if Manus today does not flow anything too sensitive, a combination of poor supervision, powerful automation and questionable data information makes it far more risky than your average AI helper.

We reached for Manus AI for commenting on this article, but we didn’t answer before our deadline.

Genai, the future of fraud and why are you perhaps an easy target

So, how can you protect yourself?

As I mentioned, if you use Manus AI or not, your personal information will continue to end in your data sets, so the best way to protect yourself to make you difficult to find information. Here are some things you can do.

1) Be careful in the use of manus: With this level of autonomy, especially since it is a very new AI (not to mention inconsistency in my privacy policy regarding management), I would not yet recommend using it. It is simply too much damage to which it can do if it does not work or subject to a rapid injection attack.

2) Keep your personal information under the lining: With so many companies selling and trading personal information, many of them with technological companies for training AI, it is a good idea to refuse your information unless you can absolutely avoid sharing it.

3) Actively remove your personal information from the Internet so that your data can not scrape modes like Manus: Set your social media to private, limit data collection through privacy settings and remove your data from the Broker data and the people’s search sites. Autonomous agents based on a cloud such as Manus are designed to draw and analyze enormous quantities of internet data, and your publicly available information can be pushed into their training data without your knowledge or consent.

Although no service promises to remove all your information from the Internet, the removal service is great if you want to constantly monitor and automate the process of removing your data from hundreds of sites continuously over a long period of time. View my top selection here for data removal services.

4) Use strong antivirus software to help defend against the wider surface of the attack created by autonomous AI systems: Manus’s ability to act independently and communicate with external systems increases the risk of opponent attacks and transmission of malicious software, so the robust safety of the endpoint is crucial to protect your devices and data.

The best way to protect yourself from malicious connections installed by malicious software, potentially access to private data, is to install antiviral software on all your devices. This protection can also alert you of phishing e -Mail and fraud from Ransomware, keeping your personal information and digital assets. Get my elections for the best winners to protect against antivirus 2025 for your Windows, Mac, Android and iOS devices.

5) Use products and services aimed at privacy: Whenever you cannot avoid sharing personal information, opt for companies that give your privacy priority. For example, use a Duckduckgo browser instead of GMALILA instead of Chrome and locks or proton mail. Check the best alternatives of private and safe browser to Google.

6) Inform all software, including your operating system, browsers and security tools, to patch vulnerability that attackers can use: As AI agents like manus act in a cloud and can communicate with your systems, separated software can become a simple goal for exploitation or unauthorized approach. Therefore, it is always a good idea to keep software, operating systems, browsers and security tools Updated at any time.

7) Enable Multifactor authentication (MFA) in all your accounts: As the attacks on Ai-Volited and automated data extraction become more frequent, MFA Adds a key layer of defense, which makes it much more difficult for the attackers to compromise the account, even if the AI ​​system leaks or scratches your credentials.

8) Use strong, unique passwords for each account: Since AI agents based on clouds like manus can communicate with external systems and potentially access sensitive data, unique passwords help prevent a violation of one account to lead to a wider compromise. Consider using password manager to generate and store complex passwords. Get more details about my Best reviewed password managers from 2025. Here.

Kurts key endeavor

As we go forward with AI like manus, it’s exciting to think about opportunities, but it’s also important to keep your eyes open about what it means for our personal information. Even if you never use manus yourself, your data could still end up in training data.

So what can you do? Be informed, download your digital print and select the tools of your privacy priority. As Ai continues to develop, it is proactive about your data is the best way to enjoy the benefits of innovation while keeping your personal information at the same time.

Do you think that companies behind powerful AI agents like Manus should follow the stricter privacy rules and independent supervision before they are allowed to collect and use people’s data? Let us know by writing us on Cyberguy.com/contact.

For more of my technological tips and security warnings, subscribe to my free newsletter about Cyberguy Report, moving toward Cyberguy.com/newslettter.

Ask the Kurt question or let us know which stories would you like to cover.

Follow Kurt on his social channels:

Answer to the most demanding questions about Cyberguy:

New from Kurt:

Copyright 2025 Cyberguy.com. All rights reserved.