THREAT OPS › Threat News › GTIG AI Threat Tracker: Adversaries Leverage AI for Vulnerability Exploitation, Augmented Operations, and Initial Access
GTIG AI Threat Tracker: Adversaries Leverage AI for Vulnerability Exploitation, Augmented Operations, and Initial Access
<div class="block-paragraph_advanced"><h3><span style="vertical-align: baseline;">Executive Summary</span></h3> <p><span style="vertical-align: baseline;">Since our </span><a href="https://cloud.google.com/blog/topics/threat-intelligence/distillation-experimentation-integration-ai-adversarial-use"><span style="text-decoration: underline; vertical-align: baseline;">February 2026 report</span></a><s
Attributed threat actors
- Mustang PandaG0129
- Threat Group-3390G0027
MITRE ATT&CK techniques
- Acquire InfrastructureT1583
- Gather Victim Host InformationT1592
- Artificial IntelligenceT1588.007
- IP AddressesT1590.005
- MalwareT1587.001
- HardwareT1592.001
- VulnerabilitiesT1588.006
- Supply Chain CompromiseT1195
- Social EngineeringT1684
- MasqueradingT1036
- Junk Code InsertionT1027.016
- Business RelationshipsT1591.002
- Gather Victim Org InformationT1591
- ToolT1588.002
- PhishingT1566
- Multi-hop ProxyT1090.003
- Obfuscated Files or InformationT1027
- Polymorphic CodeT1027.014
- ExploitsT1587.004
- CredentialsT1589.001
- Identify RolesT1591.004
- Establish AccountsT1585
- Obtain CapabilitiesT1588
- Develop CapabilitiesT1587
- ExploitsT1588.005
- Acquire InfrastructureAML.T0008
- AI Development WorkspacesAML.T0008.000
- AI Service ProxiesAML.T0008.005
- AI Supply Chain CompromiseAML.T0010
- AI SoftwareAML.T0010.001
- Obtain CapabilitiesAML.T0016
- Software ToolsAML.T0016.001
- Generative AIAML.T0016.002
- Develop CapabilitiesAML.T0017
- Establish AccountsAML.T0021
- AI Model Inference API AccessAML.T0040
- LLM JailbreakAML.T0054
- Reverse ShellAML.T0072
- MasqueradingAML.T0074
- Generate DeepfakesAML.T0088
- Generate Malicious CommandsAML.T0102
- Deploy AI AgentAML.T0103
Indicators of compromise
- eb0687daed29f3651c61b0a2aa4a0cdcf2049a1ebae2e15e2dd9326471d318a1sha256
- https://saif.google/secure-ai-framework/risksurl
- https://deepmind.google/blog/introducing-codemender-an-ai-agent-for-code-security/url
- https://www.virustotal.com/gui/file/eb0687daed29f3651c61b0a2aa4a0cdcf2049a1ebae2e15e2dd9326471d318a1url
- https://www.virustotal.com/gui/collection/malware--77a0b844-02bb-563c-b8fd-304f93e11ef1/iocsurl
- https://www.virustotal.com/gui/collection/malware--30f26e32-0393-5023-92ef-f677f1def61c/iocsurl
- https://www.virustotal.com/gui/collection/malware--6cae6e39-72de-5b9e-aebe-47243e3dc63a/iocsurl
- https://blog.virustotal.com/2026/02/from-automation-to-infection-how.htmlurl
- https://ai.google/principles/#our-ai-principles-in-actionurl
- https://gemini.google/us/policy-guidelines/?hl=enurl
- https://transparency.google/our-approach/our-policy-process/url
- https://security.googleblog.com/2025/01/how-we-estimate-risk-from-prompt.html?m=1url
- https://ai.google.dev/url
- https://ai.google.dev/responsibleurl
- https://ai.google.dev/responsible/docs/safeguardsurl
- https://ai.google.dev/responsible/docs/evaluation#red-teamingurl
- https://security.googleblog.com/2025/06/mitigating-prompt-injection-attacks.htmlurl