Chinese Hackers Used AI Agents to Breach ConnectWise Customers
Microsoft says a China state-aligned group stole cloud backup credentials and reached about 2,250 organizations while ConnectWise paused its backup service
By AXY AI Desk ·
Key points
- Microsoft says China state-aligned hackers stole application credentials from ConnectWise's cloud backup service.
- The stolen credentials could let attackers sign in to customers' Microsoft-hosted environments and reach their emails and files.
- Microsoft estimates the AI-driven campaign reached around 2,250 organizations, at almost 3,000 attempted sign-ins per minute.
- ConnectWise has paused access to its cloud backup system, is investigating, and said it was working to restore the service as of Thursday.
Chinese state-aligned hackers stole login credentials from IT company ConnectWise and used them to spy on thousands of its customers, according to security researchers at Microsoft. The attackers used AI agents, which are programs that can carry out a series of tasks with little human direction, to move quickly from one target to the next. Microsoft says the campaign reached around 2,250 organizations at a rate of almost 3,000 attempted sign-ins per minute.
For the companies caught up in it, this is not an abstract hacking story. Someone outside the business may have been able to open its email and read its files while looking like a normal, authorized user. The break-in started at a supplier, and that is why thousands of organizations that never dealt directly with the attackers are now affected.
How Stolen Keys Opened Customer Inboxes
ConnectWise makes software that IT departments and service providers use to manage computers and networks and to connect to them remotely. Many of its corporate customers also rely on its cloud backup service, which keeps copies of company data so it can be restored after a system crash, a ransomware attack or an accidental deletion.
Microsoft's researchers said the hackers broke into that cloud backup service and stole application credentials. These are digital keys that software uses to prove it is allowed to connect. In this case, the credentials establish the data links between ConnectWise and its corporate customers.
The danger lies in what those keys allow. According to Microsoft, the hackers could use them to log in to a customer's Microsoft-hosted environment as if they were an authorized user. From there, they could access that company's emails and other files. They would not need to guess a password or trick a single employee into clicking a link.
That is what makes stolen credentials so valuable to an attacker. A single break-in to a widely used service can open many doors at once, because the same kinds of keys work for every customer that depends on that service. The attackers did not have to break into each company separately. They only had to find the link that connected them all.
Why AI Agents Made the Attack Faster
Microsoft said the hackers used AI agents to work through their targets quickly. The company estimated that the campaign reached around 2,250 organizations, at a rate of almost 3,000 attempted sign-ins per minute. Those figures are the researchers' own estimates of the campaign, and they show how much ground an automated approach can cover in a short time.
Speed is the core problem. Most security teams depend on spotting unusual activity and then responding to it, usually with a person reviewing alerts and deciding what to do. Microsoft said the pace of this attack gives defenders "little time to detect and respond."
This is also why the use of AI matters beyond this one case. A group that can run thousands of attempts in parallel does not need to be lucky. It only needs a weak point somewhere in the chain, and automation makes it far more likely that one will be found.
ARTEX and the South Korea Compromise
The tool named in the research is ARTEX, which Microsoft describes as an open-source AI framework. An open-source framework is software whose code is publicly available, so that anyone can use it, study it or change it. ARTEX works with AI models to stress-test cyber defenses, meaning it is built to probe a network and find its weaknesses before a real attacker does.
That design is useful to defenders, but the same capability can be turned around. Testing tools that find gaps in a network can also be used to exploit those gaps. According to the source material, that is how the hackers used the software in this campaign.
The same AI software was used in a recent compromise of South Korea's financial systems. The researchers said many of those systems had weak security and lacked multi-factor authentication. That is a second check, such as a code sent to a phone or an approval prompt, which stops an attacker who has only a stolen password from getting in. Without that second check, a single leaked credential can be enough to get through the front door.
The overlap between the two cases is the point Microsoft's researchers appear to be making. The same tool was used against a national financial system and against a set of IT customers, and in both cases the weak spots were the kind that basic protections are meant to close.
What ConnectWise Has Done Since the Breach
ConnectWise said it has paused access to its cloud backup system and is investigating the breach. As of Thursday, the company said it was working on restoring the service.
The pause matters to customers in practical terms. Companies that depend on the backup service may be unable to reach copies of their data while the investigation continues. For a business that has just found out its supplier was compromised, that means both a security problem and an operational one, because the systems meant to recover from a disaster are now offline.
Restoring the service safely will require more than switching it back on. The stolen credentials need to be dealt with, so that they cannot be used again, and the company needs to be confident that the attackers no longer have a way back in. Restoring too early could let the same keys be used again.
Why Supplier Breaches Reach So Far
The ConnectWise case shows how a supplier can become a route into many businesses at once. A company that manages IT for other organizations often holds keys to their systems, because that is how it does its job. If those keys are stolen, the damage can spread to every client that trusted the supplier with access.
For ordinary staff, the consequences can be concrete. Email accounts can be used to read contract discussions, payroll messages or customer details. An attacker who can read real emails can also write convincing messages to colleagues and clients, asking for payments or information. Stored files may include financial records, personal data or confidential plans that were never meant to leave the company.
The case also shows how the pace of attacks is changing. Weak protections, such as passwords without a second check, become much more dangerous when an attacker can test many accounts in a short time. A password that might have held against a few guesses by hand can fail quickly when the guessing is automated.
This could push more companies to require multi-factor authentication across all their systems, and to review which outside services hold keys to their data. It could also make companies ask harder questions of their IT providers about how those keys are stored and how quickly a compromise would be detected.
Whether the Backup Service Returns Safely
The immediate question is when the cloud backup service returns, and whether it comes back with new safeguards in place. Customers will be looking for a clear account of which credentials were taken, how long the attackers had access, and whether any data was copied or changed. Those answers will determine how much trust businesses are willing to place in the service again.
Businesses that use ConnectWise may need to take their own steps. Standard practice after a credential theft includes replacing the connection keys, checking Microsoft-hosted environments for unusual sign-ins, and confirming that multi-factor authentication is switched on for every account. These steps are routine, but they take time, and the speed Microsoft described means that waiting carries its own risk.
The attribution to a China state-aligned group is Microsoft's conclusion, and it will shape how governments and companies respond. Any further detail will come from the technical findings Microsoft publishes and from whatever ConnectWise says once its investigation is complete. For now, the clearest facts are the ones the researchers have given: stolen credentials, a backup service paused for investigation, and an estimated 2,250 organizations in the crosshairs.
Questions and answers
Who is behind the ConnectWise breach?
Microsoft's security researchers say the attack was carried out by China state-aligned hackers. The researchers reached that conclusion in their analysis of the campaign.
How many organizations were affected by the ConnectWise hack?
Microsoft estimates the hackers used AI agents to reach around 2,250 organizations, at a rate of almost 3,000 attempted sign-ins per minute.
What did the hackers steal from ConnectWise?
The hackers stole application credentials from ConnectWise's cloud backup service. These credentials link ConnectWise to its corporate customers and could be used to sign in to customers' Microsoft-hosted environments.
Is ConnectWise's cloud backup service still paused?
ConnectWise said it has paused access to its cloud backup system while it investigates the breach. As of Thursday, the company said it was working on restoring the service.
Topics: ConnectWise, Microsoft, Chinese hackers, AI agents, cyberattack, cloud backup, credential theft, ARTEX
Sources: TechCrunch AI
https://axynewsglobal.com/article/chinese-hackers-used-ai-agents-to-breach-connectwise-customers