Latest
Wiz reports an open DeepSeek database[7]
Wiz said a publicly accessible ClickHouse database with no authentication held chat history and secret keys. DeepSeek secured it promptly after disclosure.
Timeline
AI data breach timeline
A record of the documented cases and the NCSC guidance published around them. This is a pattern of incidents, not one event.
Last reviewed against legislation.gov.uk, National Cyber Security Centre and OWASP GenAI Security Project. Sources
How to read it
Dates and sources
Each entry is dated from its source. Where a source gave a date for an event that happened earlier, such as the first use of the Microsoft storage link in 2020, the entry uses the earlier date [8].
The cases are different in cause. The ChatGPT incident came from a software library bug [9], the Samsung incident from staff behaviour [10], and the Microsoft and DeepSeek exposures from storage access [8] [7]. The guidance entries show how UK advice developed alongside them [3] [4].
Newer cases will be added when they are documented by the company involved or a primary source.
Timeline
How it unfolded
Newest first. Each entry links to its source.
-
-
NCSC publishes guidance for leaders on AI and cyber security[3]
The guidance tells organisations to have a response plan for AI-related incidents and to understand their supply chains.
-
NCSC publishes secure AI development guidelines[4]
Version 1.0 covers secure design, development, deployment and operation, and lists incident management processes under secure deployment.
-
Wiz publishes its findings on 38 TB of exposed data[8]
Wiz said the link gave full control of a storage account holding disk backups, secret keys and over 30,000 internal Teams messages.
-
Wiz reports the Microsoft storage exposure[8]
Wiz reported the issue to Microsoft's security response centre; Microsoft invalidated the link two days later.
-
Samsung restricts generative AI tools[10]
Reporting on a company memo said engineers had uploaded internal source code to ChatGPT in April, and that Samsung was temporarily restricting generative AI while it prepared security measures.
-
OpenAI names an open-source library bug as the cause[9]
OpenAI said a bug in the redis-py library exposed payment-related details of about 1.2% of ChatGPT Plus subscribers active in a nine-hour window, and that it was contacting those affected.
-
ChatGPT shows some users other people's chat titles[9]
OpenAI took ChatGPT offline after some users saw other users' chat queries and some Plus subscribers saw other users' email addresses.
-
Storage link first committed to GitHub[8]
Wiz said the shared link that later exposed Microsoft AI research data was first committed to a public repository on this date.
Keep reading
Read next
Questions
Sourced answers on reporting an AI data breach, the 72 hour rule, telling people, supplier breaches, leaked keys and what UK organisations should do.
Read the guide
First response
What to do first when data has leaked through an AI tool: record awareness, stop the exposure, replace secrets, scope the data and decide on reporting.
Read the guide
Response plan
What to add to your incident response plan for AI tools: owners, data stores, supplier contacts, decision rights and the NCSC's advice for leaders.
Read the guideHow CyPro can help
Need a hand working through this?
Finding out whether an incident reaches you is one job. Checking your own systems, tightening access and briefing the board are several more. CyPro's responders and consultants can take any of them on.
Incident response
If you may already be compromised, our responders contain it, preserve the evidence, find out what was touched and get you running again safely.
More on cypro.co.uk
24/7 security monitoring
Our SOC analysts watch your systems day and night for the signs of an attack in progress: unexpected processes, new outbound connections, logins that do not fit.
More on cypro.co.uk
Fractional CISO for the board
A senior security lead who advises the board on what to fix first, what to accept and what to tell customers, and owns the decision record.
More on cypro.co.uk
Penetration testing
A focused test of your systems and the controls around them, so you know they hold before an attacker finds out for you.
More on cypro.co.uk
Hardening and segmentation
Our consultants put the controls in place that limit the damage when something is missed: firewall rules, network isolation and tighter admin access.
More on cypro.co.uk
Vulnerability scanning
Find every internet-facing system you run, including the forgotten ones, and see which are still on a vulnerable build.
See the scanning service
Talk to a specialist
Worried this affects you?
When data has leaked through an AI tool, its logs or its storage, the early decisions shape what you must report and to whom. An incident response team helps you contain the exposure, establish what was involved and prepare your notifications.