Darktrace’s Mid-Year 2026 Report Finds Attackers Exploiting Trusted Identities, Services and AI
Summary
Darktrace’s H1 2026 report says attackers increasingly abused trusted identities, SaaS, cloud, supply chains and AI. Its customer data includes phishing and campaign trends, while investigations document infostealer activity and AI infrastructure attacks.
Key points
- Darktrace describes a shift toward abusing trusted accounts, SaaS platforms, cloud entitlements, supply chains, remote administration tools and non-human identities.
- In aggregated email data for H1 2026, 67% of phishing emails passed DMARC; 25.8% targeted VIP users, and 39% used novel social engineering techniques.
- Information stealers including StealC and AMOS were prevalent; ClickFix remained a common delivery route, and one campaign affected customers in 17 countries.
- The report cites supply-chain campaigns involving Axios, Trivy and Hola VPN, as well as infostealer distribution through legitimate blockchain infrastructure.
- A compromised SaaS account triggered activity across email, SaaS and network layers; Darktrace says the combined signals revealed an intrusion.
- AI service detections per deployment rose 13% to more than 16 million. In a sample of nearly 280,000 prompts, about 1% contained sensitive data.
- Darktrace reports that a React2Shell honeypot was compromised in under two hours and describes attacks on AI infrastructure, alongside state-linked activity focused on persistent access.
Article Details
- Publisher
- Darktrace
- Report Period
- 2026-01-01 to 2026-06-30
- Scope
- Threat activity and aggregated observations across Darktrace customer deployments during the first half of 2026.
- Key Statistics
- 67% of phishing emails in aggregated Darktrace / EMAIL data passed DMARC.
- 37% of phishing emails contained a high volume of text, compared with 32% in the first half of 2025; 39% featured novel social engineering techniques.
- AI service detections per deployment rose 13% in the first half of 2026, surpassing 16 million detections; the typical organization interacted with seven AI providers.
- Across nearly 280,000 prompts submitted by almost 28,000 users over 28 days, approximately 1% (2,945 instances) contained sensitive data; prompt data was analyzed in aggregate and anonymized form.
- Campaign clusters in the first half of 2026: EMEA accounted for 60% of cases, the Americas 30%; within those regions, the United States accounted for 60% of AMS clusters, and Japan for 40% of APJ customer cases.
- Recommendations
- Treat AI gateways, proxies, and model endpoints as production cloud workloads and secure and monitor them alongside identities, applications, and cloud infrastructure.
- Assess trusted activity in its wider context and identify behavioral deviations rather than relying on authentication, reputation, and provenance alone.
People
Malware
Atomic macOS Stealer (AMOS)customers were also among the most identity-centric: information stealers, with dedicated StealC and AMOS campaigns running through the half-year. Their prevalence is, fundamentally, an identity story.Phexiaattackers abused legitimate blockchain infrastructure to distribute infostealers, including AMOS and Phexia. Popular tools like VPNs, often used by users with limited security resources, combined with legitimateStealCDarktrace customers were also among the most identity-centric: information stealers, with dedicated StealC and AMOS campaigns running through the half-year. Their prevalence is, fundamentally, an identity story.
Products
Autonomous Responseare promptly reported to the relevant customer security teams. When a customer has Darktrace’s Autonomous Response technology enabled, these threats are swiftly mitigated to prevent escalation.Cyber AI Analysttoward additional AI services; when that failed, the attacker pivoted to cryptomining. Darktrace’s Cyber AI Analyst pieced the intrusion together and Darktrace’s Managed Threat Detection service alerted the customer,Darktrace / EMAILStatistics related to email are derived from aggregated Darktrace / EMAIL data across all cloud-hosted customer deployments between January 1 and June 30, 2026. Standard data-quality filtering was applied to excludeDarktrace / SECURE AIDarktrace / SECURE AI observations suggest the most common real-world risk is quieter still: employees entering personal identifiable information (PII), tax records, identity documents, company financial data, HR
Countries
Chinathrough legitimate services, cloud infrastructure, and trusted relationships. Operations linked to China, Russia, Iran, and the Democratic People’s Republic of Korea (DPRK) shared a common characteristic: aDemocratic People’s Republic of Korea (DPRK)and trusted relationships. Operations linked to China, Russia, Iran, and the Democratic People’s Republic of Korea (DPRK) shared a common characteristic: a focus on persistence and strategic positioningIranlegitimate services, cloud infrastructure, and trusted relationships. Operations linked to China, Russia, Iran, and the Democratic People’s Republic of Korea (DPRK) shared a common characteristic: a focus onJapanThe Asia-Pacific and Japan (APJ) region was less affected by campaign clusters, potentially indicating that threat actors placed a lower priority on the region and instead focused their efforts elsewhere.Russiathrough legitimate services, cloud infrastructure, and trusted relationships. Operations linked to China, Russia, Iran, and the Democratic People’s Republic of Korea (DPRK) shared a common characteristic: a focus onUkraineRussia: Darktrace investigations, alongside wider industry reporting, highlighted Russian reliance on trusted relationships and supply-chain targeting for long-term intelligence on Ukraine related support [2].United KingdomIn EMEA, the United Kingdom and Zimbabwe each accounted for 23% of identified cases, both being involved in a variety of campaign types.United Statesdistribution route. One recent campaign impacted Darktrace customers across 17 countries, with the United States the most affected. The compromise did not begin with a software flaw, but with a trusted user taking aZimbabweIn EMEA, the United Kingdom and Zimbabwe each accounted for 23% of identified cases, both being involved in a variety of campaign types.
Industries
Educationinfluenced by the distribution of Darktrace's customer base. For example, Finance, Manufacturing, and Education are strongly represented among Darktrace customers, which may result in a higher number of observedFinancial Insurance activitiesIn contrast, AMS targeting was more evenly distributed, with the Education, Public administration and defence, and Financial Insurance activities sectors all forming over 20% of AMS regional cases.Information and communicationIn EMEA, the Information and Communication was the most affected by a significant margin, representing 25% of all cases.Manufacturingthey are also influenced by the distribution of Darktrace's customer base. For example, Finance, Manufacturing, and Education are strongly represented among Darktrace customers, which may result in a higher numberPublic administration and defenceIn contrast, AMS targeting was more evenly distributed, with the Education, Public administration and defence, and Financial Insurance activities sectors all forming over 20% of AMS regional cases.