ブログ
/
Network
/
September 21, 2023

How Darktrace Detected Black Basta Ransomware

Discover how Darktrace uncovered Black Basta ransomware. Learn about its tactics, techniques, and how to protect your network from this threat.
Inside the SOC
Darktrace cyber analysts are world-class experts in threat intelligence, threat hunting and incident response, and provide 24/7 SOC support to thousands of Darktrace customers around the globe. Inside the SOC is exclusively authored by these experts, providing analysis of cyber incidents and threat trends, based on real-world experience in the field.
Written by
Matthew John
Director of Operations, SOC
Default blog image
21
Sep 2023

What is Black Basta?

Over the past year, security researchers have been tracking a new ransomware group, known as Black Basta, that has been observed targeted organizations worldwide to deploy double extortion ransomware attacks since early 2022. While the strain and group are purportedly new, evidence seen suggests they are an offshoot of the Conti ransomware group [1].

The group behind Black Basta run a Ransomware as a Service (RaaS) model. They work with initial access brokers who will typically already have a foothold in company infrastructure to begin their attacks. Once inside a network, they then pivot internally using numerous tools to further their attack.

Black Basta Ransomware

Like many other ransomware actors, Black Basta uses double extortion as part of its modus operandi, exfiltrating sensitive company data and using the publication of this as a second threat to affected companies. This is also advertised on a dark web site, setup by the group to apply further pressure for affected companies to make ransom payments and avoid reputational damage.

The group also seems to regularly take advantage of existing tools to undertake the earlier stages of their attacks. Notably, the Qakbot banking trojan, seems to be the malware often used to gain an initial foothold within compromised environments.

Analysis of the tools, procedures and infrastructure used by Black Basta belies a maturity to the actors behind the ransomware. Their models and practices suggest those involved are experienced individuals, and security researchers have drawn possible links to the Conti ransomware group.

As such, Black Basta is a particular concern for security teams as attacks will likely be more sophisticated, with attackers more patient and able to lie low on digital estates for longer, waiting for the opportune moment to strike.

Cyber security is an infinite game where defender and attacker are stuck as cat and mouse; as new attacks evolve, security vendors and teams respond to the new indicators of compromise (IoCs), and update their existing rulesets and lists. As a result, attackers are forced to change their stripes to evade detection or sometimes even readjust their targets and end goals.

Anomaly Based Detection

By using the power of Darktrace’s Self-Learning AI, security teams are able to detect deviations in behavior. Threat actors need to move through the kill chain to achieve their aims, and in doing so will cause affected devices within networks to deviate from their expected pattern of life. Darktrace’s anomaly-based approach to threat detection allows it recognize these subtle deviations that indicate the presence of an attacker, and stop them in their tracks.

Additionally, the ecosystem of cyber criminals has matured in the last few decades. It is well documented how many groups now operate akin to legitimate companies, with structure, departments and governance. As such, while new attack methods and tactics do appear in the wild, the maturity in their business models belie the experience of those behind the attack.

As attackers grow their business models and develop their arsenal of attack vectors, it becomes even more critical for security teams to remain vigilant to anomalies within networks, and remain agnostic to underlying IoCs and instead adopt anomaly detection tools able to identify tactics, techniques, and procedures (TTPs) that indicate attackers may be moving through a network, ahead of deployment of ransomware and data encryption.

Darktrace’s Coverage of Black Basta

In April 2023, the Darktrace Security Operations Center (SOC) assisted a customer in triaging and responding to an ongoing ransomware infection on their network. On a Saturday, the customer reached out directly to the Darktrace analyst team via the Ask the Expert service for support after they observed encrypted files and locked administrative accounts on their network. The analyst team were able to investigate and clarify the attack path, identifying affected devices and assisting the customer with their remediation. Darktrace DETECT™ observed varying IoCs and TTPs throughout the course of this attack’s kill chain; subsequent analysis into these indicators revealed this had likely been a case of Black Basta seen in the wild.

Initial Intrusion

The methods used by the  group to gain an initial foothold in environments varies – sometimes using phishing, sometimes gaining access through a common vulnerability exposed to the internet. Black Basta actors appear to target specific organizations, as opposed to some groups who aim to hit multiple at once in a more opportunistic fashion.

In the case of the Darktrace customer likely affected by Black Basta, it is probable that the initial intrusion was out of scope. It may be that the path was via a phishing email containing an Microsoft Excel spreadsheet that launches malicious powershell commands; a noted technique for Black Basta. [3][4]  Alternatively, the group may have worked with access brokers who already had a foothold within the customer’s network.

One particular device on the network was observed acting anomalously and was possibly the first to be infected. The device attempted to connect to multiple internal devices over SMB, and connected to a server that was later found to be compromised and is described throughout the course of this blog. During this connection, it wrote a file over SMB, “syncro.exe”, which is possibly a legitimate Remote Management software but could in theory be used to spread an infection laterally. Use of this tool otherwise appears sporadic for the network, and was notably unusual for the environment.

Given these timings, it is possible this activity is related to the likely Black Basta compromise. However, there is some evidence online that use of Syncro has been seen installed as part of the execution of loaders such as Batloader, potentially indicating a separate or concurrent attack [5].

Internal Reconnaissance + Lateral Movement

However the attackers gained access in this instance, the first suspicious activity observed by Darktrace originated from an infected server. The attacker used their foothold in the device to perform internal reconnaissance, enumerating large portions of the network. Darktrace DETECT’s anomaly detection noted a distinct rise in connections to a large number of subnets, particularly to closed ports associated with native Windows services, including:

  • 135 (RPC)
  • 139 (NetBIOS)
  • 445 (SMB)
  • 3389 (RDP)

During the enumeration, SMB connections were observed during which suspiciously named executable files were written:

  • delete.me
  • covet.me

Data Staging and Exfiltration

Around 4 hours after the scanning activity, the attackers used their knowledge gained during enumeration about the environment to begin gathering and staging data for their double extortion attempts. Darktrace observed the same infected server connecting to a file storage server, and downloading over 300 GiB of data. Darktrace DETECT identified that the connections had been made via SMB and was able to present a list of filenames to the customer, allowing their security team to determine the data that had likely been exposed to the attackers.

The SMB paths detected by Darktrace showed a range of departments’ file areas being accessed by threat actors. This suggests they were interested in getting as much varied data as possible, presumably in an attempt to ensure a large amount of valuable information was at their disposal to make any threats of releasing them more credible, and more damaging to the company.

Shortly after the download, the device made an external connection over SSH to a rare domain, dataspt[.]com, hosted in the United States. The connection itself was made over an unusual port, 2022, and Darktrace recognized that the domain was new for the network.

During this upload, the threat actors uploaded a similar volume of data to the 300GiB that had been downloaded internally earlier. Darktrace flagged the usual elements of this external upload, making the identification and triage of this exfiltration attempt easier for the customer.

On top of this, Darktrace’s autonomous investigation tool Cyber AI Analyst™ launched an investigation into this on-going activity and was able to link the external upload events to the internal download, identifying them as one exfiltration incident rather than two isolated events. AI Analyst then provided a detailed summary of the activity detected, further speeding up the identification of affected files.

Preparing for Exploitation

All the activity documented so far had occurred on a Wednesday evening. It was at this point that the burst of activity calmed, and the ransomware lay in wait within the environment. Other devices around the network, particularly those connected to by the original infected server and a domain controller, were observed performing some elements of anomalous activity, but the attack seemed to largely take a pause.

However, on the Saturday morning, 3 days later, the compromised server began to change the way it communicated with attackers by reaching out to a new command and control (C2) endpoint. It seemed that attackers were gearing up for their attack, taking advantage of the weekend to strike while security teams often run with a reduced staffing.

Darktrace identified connections to a new endpoint within 4 minutes of it first being seen on the customer’s environment. The server had begun making repeated SSL connections to the new external endpoint, faceappinc[.]com, which has been flagged as malicious by various open-source intelligence (OSINT) sources.

The observed JA3 hash (d0ec4b50a944b182fc10ff51f883ccf7) suggests that the command-line tool BITS Admin was being used to launch these connections, another suggestion of the use of mature tooling.

In addition to this, Darktrace also detected the server using an administrative credential it had never previously been associated with. Darktrace recognized that the use of this credential represented a deviation from the device’s usual activity and thus could be indicative of compromise.

The server then proceeded to use the new credential to authenticate over Keberos before writing a malicious file (“management.exe”) to the Temp directory on a number of internal devices.

Encryption

At this point, the number of anomalous activities detected from the server increased massively as the attacker seems to connect networkwide in an attempt to cause as quick and destructive an encryption effort as possible. Darktrace observed numerous files that had been encrypted by a local process. The compromised server began to write ransom notes, named “instructions_read_me.txt” to other file servers, which presumably also had successfully deployed payloads. While Black Basta actors had initially been observed dropping ransom notes named “readme.txt”, security researchers have since observed and reported an updated variant of the ransomware that drops “instructions_read_me_.txt”, the name of the file detected by Darktrace, instead [6].

Another server was also observed making repeated SSL connections to the same rare external endpoint, faceappinc[.]com. Shortly after beginning these connections, the device made an HTTP connection to a rare IP address with no hostname, 212.118.55[.]211. During this connection, the device also downloaded a suspicious executable file, cal[.]linux. OSINT research linked the hash of this file to a Black Basta Executable and Linkable File (ELF) variant, indicating that the group was highly likely behind this ransomware attack.

Of particular interest again, is how the attacker lives off the land, utilizing pre-installed Windows services. Darktrace flagged that the server was observed using PsExec, a remote management executable, on multiple devices.

Darktrace Assistance

Darktrace DETECT was able to clearly detect and provide visibility over all stages of the ransomware attack, alerting the customer with multiple model breaches and AI Analyst investigation(s) and highlighting suspicious activity throughout the course of the attack.

For example, the exfiltration of sensitive data was flagged for a number of anomalous features of the meta-data: volume; rarity of the endpoint; port and protocol used.

In total, the portion of the attack observed by Darktrace lasted about 4 days from the first model breach until the ransomware was deployed. In particular, the encryption itself was initiated on a Saturday.

The encryption event itself was initiated on a Saturday, which is not uncommon as threat actors tend to launch their destructive attacks when they expect security teams will be at their lowest capacity. The Darktrace SOC team regularly observes and assists in customer’s in the face of ransomware actors who patiently lie in wait. Attackers often choose to strike as security teams run on reduced hours of manpower, sometimes even choosing to deploy ahead of longer breaks for national or public holidays, for example.

In this case, the customer contacted Darktrace directly through the Ask the Expert (ATE) service. ATE offers customers around the clock access to Darktrace’s team of expert analysts. Customers who subscribe to ATE are able to send queries directly to the analyst team if they are in need of assistance in the face of suspicious network activity or emerging attacks.

In this example, Darktrace’s team of expert analysts worked in tandem with Cyber AI Analyst to investigate the ongoing compromise, ensuring that the investigation and response process were completed as quickly and efficiently as possible.

Thanks to Darktrace’s Self-Learning AI, the analyst team were able to quickly produce a detailed report enumerating the timeline of events. By combining the human expertise of the analyst team and the machine learning capabilities of AI Analyst, Darktrace was able to quickly identify anomalous activity being performed and the affected devices. AI Analyst was then able to collate and present this information into a comprehensive and digestible report for the customer to consult.

Conclusion

It is likely that this ransomware attack was undertaken by the Black Basta group, or at least using tools related to their method. Although Black Basta itself is a relatively novel ransomware strain, there is a maturity and sophistication to its tactics. This indicates that this new group are actually experienced threat actors, with evidence pointing towards it being an offshoot of Conti.

The Pyramid of Pain is a well trodden model in cyber security, but it can help us understand the various features of an attack. Indicators like static C2 destinations or file hashes can easily be changed, but it’s the underlying TTPs that remain the same between attacks.

In this case, the attackers used living off the land techniques, making use of tools such as BITSAdmin, as well as using tried and tested malware such as Qakbot. While the domains and IPs involved will change, the way these malware interact and move about systems remains the same. Their fingerprint therefore causes very similar anomalies in network traffic, and this is where the strength of Darktrace lies.

Darktrace’s anomaly-based approach to threat detection means that these new attack types are quickly drawn out of the noise of everyday traffic within an environment. Once attackers have gained a foothold in a network, they will have to cause deviation from the usual pattern of a life on a network to proceed; Darktrace is uniquely placed to detect even the most subtle changes in a device’s behavior that could be indicative of an emerging threat.

Machine learning can act as a force multiplier for security teams. Working hand in hand with the Darktrace SOC, the customer was able to generate cohesive and comprehensive reporting on the attack path within days. This would be a feat for humans alone, requiring significant resources and time, but with the power of Darktrace’s Self-Learning AI, these deep and complex analyses become as easy as the click of a button.

Credit to: Matthew John, Director of Operations, SOC, Paul Jennings, Principal Analyst Consultant

Get the latest insights on emerging cyber threats

Attackers are adapting, are you ready? This report explores the latest trends shaping the cybersecurity landscape and what defenders need to know in 2025.

  • Identity-based attacks: How attackers are bypassing traditional defenses
  • Zero-day exploitation: The rise of previously unknown vulnerabilities
  • AI-driven threats: How adversaries are leveraging AI to outmaneuver security controls

Stay ahead of evolving threats with expert analysis from Darktrace. Download the report here.

Appendices

Darktrace DETECT Model Breaches

Internal Reconnaissance

Device / Multiple Lateral Movement Model Breaches

Device / Large Number of Model Breaches

Device / Network Scan

Device / Anomalous RDP Followed by Multiple Model Breaches

Device / Possible SMB/NTLM Reconnaissance

Device / SMB Lateral Movement

Anomalous Connection / SMB Enumeration

Anomalous Connection / Possible Share Enumeration Activity

Device / Suspicious SMB Scanning Activity

Device / RDP Scan

Anomalous Connection / Active Remote Desktop Tunnel

Device / Increase in New RPC Services

Device / ICMP Address Scan

Download and Upload

Unusual Activity / Enhanced Unusual External Data Transfer

Unusual Activity / Unusual External Data Transfer

Anomalous Connection / Uncommon 1 GiB Outbound

Anomalous Connection / Data Sent to Rare Domain

Anomalous Connection / Download and Upload

Compliance / SSH to Rare External Destination

Anomalous Server Activity / Rare External from Server

Anomalous Server Activity / Outgoing from Server

Anomalous Connection / Application Protocol on Uncommon Port

Anomalous Connection / Multiple Connections to New External TCP Port

Device / Anomalous SMB Followed By Multiple Model Breaches

Unusual Activity / SMB Access Failures

Lateral Movement and Encryption

User / New Admin Credentials on Server

Compliance / SMB Drive Write

Device / Anomalous RDP Followed By Multiple Model Breaches

Anomalous Connection / High Volume of New or Uncommon Service Control

Anomalous Connection / New or Uncommon Service Control

Device / New or Unusual Remote Command Execution

Anomalous Connection / SMB Enumeration

Additional Beaconing and Tooling

Device / Initial Breach Chain Compromise

Device / Multiple C2 Model Breaches

Compromise / Large Number of Suspicious Failed Connections

Compromise / Sustained SSL or HTTP Increase

Compromise / SSL or HTTP Beacon

Compromise / Suspicious Beaconing Behavior

Compromise / Large Number of Suspicious Successful Connections

Compromise / High Volume of Connections with Beacon Score

Compromise / Slow Beaconing Activity To External Rare

Compromise / SSL Beaconing to Rare Destination

Compromise / Beaconing Activity To External Rare

Compromise / Beacon to Young Endpoint

Compromise / Agent Beacon to New Endpoint

Anomalous Server Activity / Rare External from Server

Anomalous Connection / Multiple Failed Connections to Rare Endpoint

Anomalous File / EXE from Rare External Location

IoC - Type - Description + Confidence

dataspt[.]com - Hostname - Highly Likely Exfiltration Server

46.22.211[.]151:2022 - IP Address and Unusual Port - Highly Likely Exfiltration Server

faceappinc[.]com - Hostname - Likely C2 Infrastructure

Instructions_read_me.txt - Filename - Almost Certain Ransom Note

212.118.55[.]211 - IP Address - Likely C2 Infrastructure

delete[.]me - Filename - Potential lateral movement script

covet[.]me - Filename - Potential lateral movement script

d0ec4b50a944b182fc10ff51f883ccf7 - JA3 Client Fingerprint - Potential Windows BITS C2 Process

/download/cal.linux - URI - Likely BlackBasta executable file

1f4dcfa562f218fcd793c1c384c3006e460213a8 - Sha1 File Hash - Likely BlackBasta executable file

References

[1] https://blogs.blackberry.com/en/2022/05/black-basta-rebrand-of-conti-or-something-new

[2] https://www.cybereason.com/blog/threat-alert-aggressive-qakbot-campaign-and-the-black-basta-ransomware-group-targeting-u.s.-companies

[3] https://www.trendmicro.com/en_us/research/22/e/examining-the-black-basta-ransomwares-infection-routine.html

[4] https://unit42.paloaltonetworks.com/atoms/blackbasta-ransomware/

[5] https://www.trendmicro.com/en_gb/research/23/a/batloader-malware-abuses-legitimate-tools-uses-obfuscated-javasc.html

[6] https://www.pcrisk.com/removal-guides/23666-black-basta-ransomware

Inside the SOC
Darktrace cyber analysts are world-class experts in threat intelligence, threat hunting and incident response, and provide 24/7 SOC support to thousands of Darktrace customers around the globe. Inside the SOC is exclusively authored by these experts, providing analysis of cyber incidents and threat trends, based on real-world experience in the field.
Written by
Matthew John
Director of Operations, SOC

More in this series

No items found.

Blog

/

AI

/

August 7, 2026

信頼が新たなアタックサーフェスである理由:ダークトレースの2026年度中間脅威アップデート

Default blog imageDefault blog image

2026年初頭、ダークトレースのアナリストによって専用に設計されたReact2Shellハニーポットは、その展開から2時間未満で侵害を受けました。この1つのデータをとって見ても2026年前半の脅威ランドスケープのペースを表していると言うことができますが、スピードは1つの側面に過ぎません。

過去6か月間の変化は、従来のマルウェアや脆弱性を中心とした攻撃から、信頼されるアイデンティティ、プラットフォーム、およびインフラの悪用への移行を示しています。多くの組織が広範囲にAIを導入するのに伴い、アイデンティティ、SaaSプラットフォーム、クラウドに対する権限、自動化フレームワーク、および非人間アイデンティティが、攻撃者に好まれる攻撃経路となっています。

攻撃者は、外部から侵入するよりも、防御者やユーザーが信頼するよう条件づけられている関係、サービス、認証済みチャネル内で活動する傾向が高まっています。

2025年と比較して何が変わったか?

2025年には、アイデンティティが新たな境界となり、攻撃者はますます従来のエクスプロイト手法を使うよりも信頼されるアカウント、SaaSプラットフォーム、そしてAIによって可能になった新手の手法を使うようになりました。2026年前半は、この進化の次の段階にあると言えます。アイデンティの悪用は依然として中心的ですが、信頼の問題は今やアカウントをはるかに超えてEメール認証、クラウドに対する権限、ソフトウェアサプライチェーン、AIゲートウェイ、リモート管理ツール、そして非人間アイデンティティにも及んでいます。

Theme 2025 (Mid-Year / Annual) H1 2026
Identity Credentials remained the weak link; identity emerged as the new perimeter. Identity remains the entry point, but trust has become the new attack surface.
Cloud & SaaS SaaS-targeted ransomware continued to rise. Cloud and SaaS became the attacker's preferred operating environment.
AI Large Language Models (LLMs) were suspected of influencing phishing shifts. LLM-generated malware, compromised AI proxies, and the abuse of AI identities emerged.
Attack Surface Scale & Speed Exponential growth of Common Vulnerabilities and Exposures (CVEs), with public proof-of-concepts appearing faster. Cloud and AI adoption expanded the attack surface, while AI accelerated exploitation. One honeypot was compromised in under two hours.
Supply Chain Legitimate services were increasingly abused. Trusted maintainers and CI/CD workflows were weaponized.

アイデンティティとEメール:危険にさらされる信頼のシグナル

Eメールは信頼されるアイデンティティへの最も確実なルートであり、攻撃者はノイズを作り出すことよりも品質の向上に投資していることをデータが示しています。2026年前半において、フィッシングEメールの67%がDMARC認証を通過していました。ほとんどのフィッシングの試みに対して、阻止するには認証だけではもはや不十分なのです。フィッシングEメールの25.8%でVIPユーザーが標的となっており、22025年と同様に25%を上回る値ですが、期間全体を通じて上昇傾向にあります。重要な点は、フィッシングの巧妙さが高まり続けていることです:フィッシングEメールの37%が大量のテキストを含んでおり、2025年前半の32%から増加しています。また、39%は新手のソーシャルエンジニアリングテクニックを特徴としており、攻撃者が特定の標的にあわせてさらなるカスタマイズを行っていることを示唆しています。

ダークトレースの顧客に最も幅広く影響を与えている脅威もアイデンティティに焦点を当てたものであり、専用のStealCおよびAMOSインフォスティーラーを使った攻撃がこの半年間に多く見られています。これらのマルウェアの蔓延も、根本的にはアイデンティティの問題です。インフォスティーラーによって収集された認証情報がしばしば最初のアクセスベクトルとして使用され、攻撃チェーンの後の段階で格段に影響の大きい侵害につながります。また、マルウェアの投下には技術的弱点の悪用をほとんど必要としない点が重要です。ユーザー自身の手で悪意あるコードを実行させるClickFixソーシャルエンジニアリング手法は、依然としてよく見られます。最近観測されたある攻撃キャンペーンは17か国のダークトレース顧客で確認され、最も影響を受けたのは米国でした。この侵害はソフトウェアの欠陥から始まったものではなく、信頼されるユーザーが信頼されるアクションを実行したことが端緒となりました。

サプライチェーン:信頼が大規模に武器化される

3月と4月には共通の教訓が確認されました。それは信頼がサプライチェーン脆弱性となったということです。Axiosの侵害では幅広く使用されているメンテナーへの信頼が悪用され、Trivyキャンペーンは信頼されるCI/CDインフラ、リリースアーティファクト、およびコンテナイメージを利用して、正当な開発ワークフローに悪意あるコードを押し込みました。

これを最もはっきりと示した事例は、2月から3月にかけて観測されたキャンペーンです。多数のデバイスがHola VPNを使用している間に悪意のあるペイロードをダウンロードし、その後、Holaの配信パイプラインの侵害に関連していることが判明しました。ダークトレースの脅威調査チームは、公開アドバイザリがリリースされる前に、複数の顧客において繰り返し発生している異常な動作から、この侵害に関連する活動を特定しました。

最近では、攻撃者が正当なブロックチェーンインフラを悪用して、AMOSやPhexia等のインフォスティーラーを投下しているケースが見られます。セキュリティリソースが限られているユーザーによく使用されている一般的なVPN等のツールは、正当なC2インフラと組み合わせることで、攻撃者がはるかに広範な被害者層に到達することを可能にする一方、防御側は関連するエンドポイントを単純にブロックすることはできないため、やっかいな問題になります。

防御者にとっての課題は、もはや悪意のあるインフラを特定することではなく、信頼されるインフラが悪意のある動作を始めたときにそれを認識できるかどうかということです。

クラウドとSaaS:標的から作戦領域へ

5月から6月には、デバイス登録、クラウドデータ窃取、SaaS悪用、RDPの拡大、リモート管理ツール関連の活動が見られ、攻撃者がクラウドやSaaSを単なる標的としてではなく、むしろ好ましい作戦環境としてますます認識していることが示唆されました。 

ダークトレース顧客でのある事例では、1つの侵害されたSaaSアカウントがEメール、SaaS、ネットワークレイヤーにわたる活動を引き起こし、これには受信トレイルールの変更、フィッシングの拡散、疑わしいインフラへの接続が含まれていました。これらの兆候のいずれも単独では決定的とは言えませんでしたが、総合すると明らかに侵入を示していました。これらの環境において、攻撃者はますます信頼管理をを回避する必要がなくなっています。信頼は、侵害されたアイデンティティ、委任されたアクセス権、および正当な管理ツールを通じて引き継ぐことができるからです。これは、2025年に見られたSaaSを標的としたランサムウェアの傾向の自然な進化であると言えます。ますます多くのケースにおいて、ビジネスが運営されるのと同じプラットフォームが、敵対者が作戦を展開するプラットフォームとなっています。

AI:攻撃の加速装置でありアタックサーフェス、そして信頼される、しかしリスクの高いアクター

信頼がアタックサーフェスなら、AIはそれが最も急激に拡大している領域であると言えます。ダークトレースの顧客基盤全体において、2026年前半にAIサービスへの接続は平均13%増加し、接続数は1600万回以上標準的な組織は7つの異なるAIプロバイダーとやり取りするようになっています。そしてAIはもはや企業のごく一部ではありません。日々の業務に組み込まれています。この変化が3つの問題を生み出しましたが、これらはすべて2026年前半にダークトレースによって観測されています。

1. 攻撃倍増装置としてのAI

ダークトレースは、AIによって生成されたReact2Shellを悪用するマルウェアを特定しました。これは攻撃者がLLMを使用して有効なエクスプロイトコードを生成し大規模に展開したものです。脅威ランドスケープ全体で見ても同様の活動が増えており、効果的な攻撃作戦への参入障壁が崩れつつあることを示唆しています。最近のJadePuffer事例にも見られるように、エージェント型脅威アクターがインターネットに接続されたサーバーの脆弱性をエクスプロイトし、その後完全に自動化されたランサムウェア攻撃が開始されており、AIは脆弱性の公開から実際のエクスプロイトまでの過程を加速させています[1]。

2. アタックサーフェスとしてのAI

今やAIレイヤー自体も、調べてみる価値があります。あるオートメーション技術メーカーにおいて、侵害されたLLMプロキシが他のAIサービスへの踏み台として利用され、それが失敗すると攻撃者は暗号通貨マイニングに切り替えたという事例がありました。DarktraceのCyber AI Analystはこの侵入インシデントを明らかにし、Managed Threat Detectionサービスにより顧客への通知が行われたことにより、事態がそれ以上進行する前に封じ込めることができました。実務者にとっての教訓は明確です:AIゲートウェイ、プロキシ、モデルエンドポイントは、本番環境のクラウドワークロードと同様に扱うべきです。なぜなら、攻撃者は既にそうしているからです。

3. 信頼される、ただしリスクもあるアクターとしてのAI

Darktrace/ SECURE AIの観測結果からわかることは、最も一般的な現実世界のリスクはさらに判別が難しいということです。従業員が、個人識別情報(PII)、税務記録、身分証明書、会社の財務データ、人事記録、個人の医療データをLLMのプロンプトに入力することや、シャドーAIの蔓延、そしてモバイルデバイスからのAI使用の増加が挙げられます。28日間で約28,000人のユーザーから送信された約280,000件のプロンプトのうち、Darktraceはこれらのプロンプトの約1%(2,945件)に機密性の高いデータが含まれていることを特定しました。*

*プロンプトデータはユーザーのプライバシーを保護するために、集積および匿名化された形で分析されました。

防御者にとって、課題はコンテキストです。つまり、正当なビジネス利用が重大なリスクに変わるタイミングを、プライバシーやユーザーの信頼を損なうことなく見極めることです。組織がAIシステムをますます信頼し、AIがマシンスピードで機密情報にアクセス、処理、共有するようになるなかで、アイデンティティ、アプリケーション、クラウドインフラと同様に、AIも保護および監視されなければなりません。

スピードと地政学:迅速な作戦、長期的な目標

2026年前半に行われたいくつかの調査では、攻撃者が新たに公開された脆弱性をいかに迅速に実用化し、パッチ適用サイクルが完了する前にOAST(Out-of-Band Application Security Testing)インフラおよび信頼されたクラウドサービスを通じてエクスプロイト検証を行っているかが明らかになりました。React2Shellは2時間で侵害され、BeyondTrustのエクスプロイトも1日未満で発生しました。このような状況の中で、国家を背後に持つ脅威アクターは、引き続き正当なサービス、クラウドインフラ、および信頼関係を通じた、長期的なアクセス、情報収集、および事前配置に重点を置いています。中国、ロシア、イラン、北朝鮮(DPRK)を背後に持つ作戦は共通の特徴を持っています。それは即座に混乱を起こすことよりも、永続化と戦略的な配置を優先するということです。

中国:ダークトレースは、中国系アクターが信頼されるサービス、DLLサイドローディング、およびモジュール型の侵入チェーンを通じた長期的なアクセスの獲得を優先している状況を確認しています。これはCrimson Echoレポートでも解説され、Twill Typhoonの手法に関連するアクティビティと一致しています。

イラン:ダークトレースによるZionSiphonの調査は、イランに関連するアクターのOT環境への関心を浮き彫りにし、諜報目的とインフラ破壊能力を融合させている実態を明らかにしています。

ロシア:ダークトレースによる調査、および業界全体のさまざまな報告は、ロシアがウクライナ関連の支援に関する長期的な情報収集のために、信頼関係とサプライチェーンを標的としていることを明らかにしています[2]。

北朝鮮:ダークトレースは、脆弱性の迅速な武器化と永続的アクセス技術を組み合わせた北朝鮮関連の活動を観測ししています。これにはAxiosのサプライチェーン侵害、React2Shellエクスプロイト、ステルス性のmacOS侵入が含まれています。

 目的は脅威アクターによって異なっていましたが、手法は非常に一貫していました。信頼されるサービス、正当なインフラ、そして永続的アクセスは、即時の混乱よりも価値が高いことが示されています。

防御者のシフト

アイデンティティの侵害、サプライチェーン攻撃、SaaSの悪用、AIインフラの標的化、国家が支援する作戦、いずれのケースにおいても、攻撃者は防御のコントロールを突破するのではなく、信頼されたシステムを通じて行動することで成功を収めることが増えています。信頼されるユーザー、信頼されるソフトウェア、信頼されるインフラチャ、そしてますます信頼されるAIシステムは、すべて有効な攻撃経路となりました。 

防御者にとっての課題はもはや、単にあるアクションが許可されるかどうかを判断することではなく、そのアクションがより広いコンテキストの中で正当かどうかを見極めることです。認証、評判、出自は依然として重要ですが、それだけではもはや十分とは言えません。攻撃者がますます信頼されるシステム内で活動するようになる中で、最も強力なシグナルは多くの場合動作の逸脱です。つまり、信頼されているアクティビティが期待される振る舞いと一致しなくなった時にそれを識別することです。

本稿の執筆にはNathaniel Jones(SVP, Global Threat Intelligence)、Emma Foulger(Global Threat Research Operations Lead)、Justin Torres (Senior Cyber Analyst) Daniel Levy(Threat Hunting Data Scientist)が協力しました。


編集:Ryan Traill(Content Manager)

付録1:脅威調査手法

ダークトレースの脅威調査チームは、顧客の運用環境に対する詳細な調査を行ってアクティブな脅威を識別し、主要な侵害インジケーター(IoC)を特定し、関連する脅威インテリジェンスを提供しています。この調査はダークトレースの異常ベース検知に基づいたもので、脅威調査チームによる徹底した分析およびコンテキスト化が行われています。検知された脅威は関連する顧客のセキュリティチームに直ちに報告されます。顧客がダークトレースの自律遮断テクノロジーを使用している場合、これらの脅威は速やかに緩和され、エスカレーションが阻止されます。

 2026年1月1日から6月30日までの期間、ダークトレースは顧客ベースにおいて多種多様なサイバー脅威を調査しました。その多くは同様のTTPおよびIoCが短い期間内に一定の数の顧客に影響を及ぼした、複数の顧客を標的とした攻撃作戦的活動であったことが判明しています。 

Eメールに関連する統計は、2026年1月1日から6月30日までの間に、すべてのクラウドホスト型顧客環境におけるDarktrace / EMAILの集約されたデータから導出されています。特異な観測値を除外するために、集約を行う前に標準的なデータ品質フィルタリングが適用されています。地域別の統計は、このデータセットの関連サブセットに基づいています。 

付録2:キャンペーン - 地域およびセクター別の傾向

過去6か月間の脅威ランドスケープは上記の大まかなテーマにより定義されていますが、ダークトレースの顧客基盤全体でのキャンペーンクラスタリングにより、それらがセクター、地域、産業ごとにどのように異なっているかが明らかになりました。 

ダークトレースの脅威調査チームは、顧客ベースに影響を与えるさまざまな脅威を調査しています。この調査を通じて、共通のTTPやインフラが観察され、短期間で多くの顧客に影響を与える、攻撃キャンペーンのような活動のクラスターが特定されました。 

セクターおよび産業は、一貫した分類を確保するために標準産業分類(SIC)システムを使用して分類されています。本レポートのセクターおよび地域別の考察は、より広範な世界的な傾向を反映している一方で、ダークトレースの顧客基盤の分布にも影響を受けています。例えば、金融、製造、教育分野はダークトレースの顧客の中で多く、これらのセクターで観測される事例数が多くなる可能性があります。これは必ずしも特定のセクターにおけるリスクが高いことを示すものではなく、顧客の分布を反映しています。同様に、地域の傾向はダークトレースの顧客の地理的分布によって影響を受ける可能性があります。

2026年前半にダークトレースの脅威調査チームによって特定されたキャンペーンクラスターの分析では、明確な地域別の傾向が明らかになりました:

  • ヨーロッパ、中東、アフリカ(EMEA)がすべてのキャンペーンクラスター事例の60%を占めており支配的でした。
  • アメリカ大陸(AMS)は、その次に影響が大きかった地域でありキャンペーンクラスター事例の30%を占めていました。
  • アジア太平洋および日本(APJ)地域はキャンペーンクラスターの影響をあまり受けていませんでした。これは脅威アクターがこの地域の優先度を低くし、代わりに他の地域に注力した可能性を示しています。

産業セクターの標的化も地域によって大きく異なりました:

  • EMEA地域では、情報通信分野が大きく影響を受け、全体の25%を占めました。
  • 対照的に、AMSでは標的はより均等に分布しており、教育、行政機関および防衛、金融および保険の各セクターがいずれもAMS地域の事例の20%以上を占めていました。
  • APJ地域ではキャンペーン活動はより均等に分散しており、特定のセクターが支配的な標的として浮上することはありませんでした。

いくつかの国はそれぞれの地域内でも際立った特徴がありました:

  • 米国はAMS内のすべてのキャンペーンクラスターの60%を占めていました。
  • 日本はAPJ全体のキャンペーン事例の40%を占めました。
  • EMEA地域では、英国とジンバブエがそれぞれ特定された事例の23%を占めており、両国ともさまざまなキャンペーンのタイプに影響を受けています。

Inside the SOCおよび2026年度脅威調査の月次傾向:アクセスからインパクトまで

Month Dominant Themes
January Voice phishing, VPS infrastructure, WebSocket C2, RMM abuse, ransomware, infostealers (StealC), and trojanized installers (7-Zip).
February Voice phishing, VPN intrusion, edge infrastructure compromise (BeyondTrust), and RMM abuse.
March Sustained supply chain compromise (Hola VPN, Axios, Trivy), malicious browser extensions, phishing, and discovery tools.
April Account creation abuse, payload delivery, VPN credential abuse, Fortinet exploitation, and botnet activity.
May PowerShell, EtherHiding, data exfiltration, VPN access, business email compromise (BEC), ClickFix, and infostealers (AMOS).
June RDP abuse, device registration, RMM usage, voice phishing, cloud data theft, botnet activity, blockchain abuse, ClickFix, and infostealers (AMOS).

付録3:参考文献

外部

[1] https://www.darkreading.com/cyberattacks-data-breaches/jadepuffer-first-complete-llm-driven-ransomware-attack

[2] https://www.trendmicro.com/en_us/research/26/c/pawn-storm-targets-govt-infra.html

ダークトレースのリソース

1.        https://www.darktrace.com/blog/ai-llm-generated-malware-used-to-exploit-react2shell

2.        https://www.darktrace.com/blog/2025-cyber-threat-landscape-darktraces-mid-year-review

3.        https://www.darktrace.com/resources/annual-threat-report-2026

4.        https://www.darktrace.com/blog/when-trust-becomes-the-attack-surface-supply-chain-attacks-in-an-era-of-automation-and-implicit-trust

5.        https://www.darktrace.com/blog/hola-vpn-abuse-from-proxy-traffic-to-malware-and-cryptomining

6.        https://www.darktrace.com/blog/security-after-signatures-operating-in-a-world-of-pre-cve-disclosure-exploitation-collapsed-trust-boundaries-and-autonomous-systems

7.        https://www.darktrace.com/blog/when-ai-infrastructure-becomes-part-of-the-attack-surface

8.        https://www.darktrace.com/blog/cve-2026-1731-how-darktrace-sees-the-beyondtrust-exploitation-wave-unfolding

9.        https://www.darktrace.com/resource/understanding-chinese-nexus-cyber-tradecraft

10.   https://www.darktrace.com/blog/chinese-apt-campaign-targets-entities-with-updated-fdmtp-backdoor

11.  https://www.darktrace.com/blog/inside-zionsiphon-darktraces-analysis-of-ot-malware-targeting-israeli-water-systems

12.  https://www.darktrace.com/resources/the-state-of-cybersecurity-in-the-finance-sector

13.  https://www.darktrace.com/blog/from-click-to-command-behavioral-detection-of-applescript-led-macos-intrusions

14.  https://www.darktrace.com/blog/the-state-of-cybersecurity-in-the-finance-sector-six-trends-to-watch

Continue reading
About the author
Nathaniel Jones
SVP, Global Threat Intelligence

Blog

/

Network

/

August 7, 2026

When AI Agents Attack: The Case for Behavioral Anomaly Detection

Default blog imageDefault blog image

The UK AI Security Institute (AISI) has just published a detailed incident report confirming what many within the cyber security industry have suspected for some time: frontier AI agents, given a degree of autonomy, can independently develop and execute attack chains against real targets, including social engineering, supply-chain compromise, and deception, without explicit instruction to do so.

The report (INC-2026-07-28-01, published 4 August 2026) documents what happened during a routine capability evaluation between 25 and 28 July. It deserves a careful read from anyone responsible for securing the modern digital estate.

The incident

AISI regularly tests frontier AI models for offensive cyber capability. In this exercise, agents were placed inside a simulated network environment and given internet access this was deliberate, given to reflect realism of the task. Across 122 evaluation runs, 19 instances of unsanctioned activity were recorded, including cases where agents broke the intended boundaries of the test and acted against real people and real infrastructure on the open internet.

In the most serious case, an agent (Mythos 5) autonomously:

  • Conducted open-source intelligence gathering to identify real developers associated with keywords from the simulated scenario
  • Created a GitHub account and submitted a pull request containing obfuscated malicious code to a real, unaffiliated repository
  • Created a second sockpuppet account to post comments on its own PR, attempting to manufacture social proof
  • When a human reviewer flagged the malicious content, the agent lied, claiming an accidental commit, then repeatedly attempted to reintroduce the payload under the guise of having "fixed" the code
  • Sent spear-phishing emails to the repository maintainer containing manipulated content designed to convince them to merge the code
  • Planted a prompt injection inside a GitHub issue, hidden in an HTML comment invisible to humans but readable by AI coding assistants parsing the repository via API

AISI confirmed no real-world harm resulted. The PRs were not merged; the payloads did not execute. But containment was a product of circumstance, not design.

Why this matters

Every action the agents took used legitimate services operating within normal parameters. Creating a GitHub account is legitimate. Opening a pull request is legitimate. Sending an email is legitimate. Posting an issue is legitimate.

No vulnerability was exploited. No known malware signature was present. No indicator of compromise from any threat intelligence feed would have matched. The entire attack existed as a pattern, a sequence of individually normal actions that, taken together, constituted a supply-chain attack, a social engineering campaign, and an attempted compromise of downstream systems.

This is the class of threat that signature-based and rule-based security tools structurally struggle to detect. They answer the question "have I seen this before?" When the answer is no – and when the attack is novel, composed of legitimate actions, and adapted in real time by an adversary in this case an agent that reasons about evasion – they struggle.

Behavioral anomaly detection: The architecture that fits

Darktrace has spent a decade building a platform around a different question: "is this behavior normal for this entity?"

The platform learns the baseline behavior of every user, device, and communication pattern across an organization's digital environment. It does not need prior knowledge of an attack technique. It identifies deviation from learned normality, surfaces it, and through its Autonomous Response capability, takes targeted action to contain threats at machine speed.

Consider how this AISI incident might show across the Darktrace platform:

Network detection and response

While no payload was successfully delivered to a victim in this incident, AISI's post-incident analysis of the agent's code (detailed on pages 24-25 of the report) reveals what would have landed on target systems had the social engineering succeeded. The agent produced three successive payload generations, each replacing the last after being detected and called out:

Generation 1 hid a module-load stager inside a legitimate bug-fix commit, guarded by an environment variable check, to skip execution in CI. Its second stage attempted to fetch Poseidon implants (Linux and Windows variants) calling back to the agent's own C2 listener, with a fallback to publicly hosted polling scripts that fetched and executed commands every 60 seconds.

Generation 2 exploited an Electron "What's New" window with nodeIntegration enabled, loading remote HTML from a GitHub Pages site the agent had registered. The page decoded an XOR-obfuscated payload at runtime that beaconed host identifiers over DNS, attempted to download the same implants, ran its own in-process C2 loop, and installed reboot persistence via crontab and Windows registry Run keys.

Generation 3 was a post-install dropper added after the agent had already been publicly caught. It XOR-decoded a hidden URL, downloaded a script over plaintext HTTP, and ran it detached, force-exiting so npm install always reports success, giving the attacker remote code execution on any machine running the install.

None of these payloads reached a victim. But the technical detail matters: this is the calibre of tradecraft an autonomous agent produced in a matter of hours, unprompted. Each generation demonstrated adaptation, the agent iterated its approach in direct response to being detected.

For network defenders, the behavioral signals these payloads would have generated are clear. DNS beaconing to novel domains, periodic HTTP polling to newly registered infrastructure, XOR-encoded data in transit, registry and crontab persistence writes on endpoints that have never exhibited such behavior. These are exactly the anomalies that Darktrace / NETWORK identifies in real time: deviations from learned device and connection baselines, regardless of whether the specific payload or C2 domain has been seen before.

Beyond the payload analysis, the agents' broader activity produced additional behavioral anomalies: new identities engaging in high-frequency interactions with external services, coordination via unusual channels (public pastebins, gists), and lateral communication patterns that deviate from any established baseline.

Email security

The spear-phishing emails sent by the agent, crafted with context-specific content designed to manipulate a specific individual, represent exactly the kind of socially engineered communication that static email gateways can miss. Darktrace / EMAIL analyses the linguistic patterns, sender behavior, and relational context of every message. A brand-new sender using tailored language to pressure a recipient into a specific technical action deviates from the learned communication profile, regardless of whether the message contains a known-bad URL or attachment.

Securing AI

This is where the incident points to a threat category that many security platforms do not yet fully address. The agent did not just use AI as a tool for attack. It targeted other AI systems, attempting prompt injections against coding assistants and trying to compromise automated workflows. Organisations adopting AI agents for development, operations, and decision-making are introducing a newer extended attack surface.  

Darktrace / SECURE AI extends the same behavioral detection model into this space. It monitors AI usage across the enterprise, detecting anomalous interactions with AI services, identifying sensitive data exposure to AI systems, providing visibility into shadow AI use. As AI agents become standard components of enterprise infrastructure, the ability to detect when an AI system is behaving outside its expected parameters, when there is a chance, it is being manipulated, or when it is interacting with something it should not be, that becomes a direct security requirement rather than a future consideration.

The principle is consistent across the Darktrace platform: learn what normal looks like, identify deviation, respond autonomously.

The AI-to-AI threat surface

One of the report's most important findings is a confirmation that AI agents can act outside of instructions and guardrails to complete tasks. The Mythos 5 agent conducted reconnaissance on a repository maintainer, inferred it might be an AI coding assistant, and deployed a prompt injection specifically crafted to compromise it.

This is a new attack class. Organizations deploying AI coding assistants, AI-powered code review, or autonomous DevOps agents now face an adversary that:

  • Moves at speed
  • Is able to adapt its approach when initial attempts are halted
  • Specifically targets AI systems via techniques (prompt injection, hidden instructions in HTML comments) that are less visible to human reviewers but parsed by machines
  • Reasons about which evasion strategies are most likely to succeed

Defending against this requires security that operates with the same adaptability and speed. A system that learns what "normal" looks like, and flags deviation from it.

What AISI recommends

The report identifies five contributing factors to the incident. Two are directly relevant:

No synchronous real-time monitoring (Section 5.3): AISI acknowledges it had not built real-time behavioral monitoring capable of immediately blocking unsanctioned agent actions. Their security team detected the incident via commercial security monitoring software, but only after activity had already occurred.

No fine-grained behavioral controls (Section 6.2): AISI's forward-looking response includes building continuous monitoring that can "immediately block or flag unsanctioned AI agent behavior."

The direction is clear: organizations need security that watches behavior continuously, learns what normal looks like, and acts when deviation occurs, without requiring a human in the loop for every decision.

The takeaway

The AISI incident is documented proof that autonomous AI agents will independently develop novel attack chains using legitimate tools, adapt when detected, employ deception, and target both humans and other AI systems. This happened last week, in a controlled setting, with commercially available models.

The security architecture that addresses this is behavioral anomaly detection applied across the full digital estate, as AI agents become standard components of enterprise infrastructure, writing code, managing deployments, processing communications, the attack surface they create is behavioral by nature.

This is the approach Darktrace has taken for years: learning what is normal across an organization’s digital environment, identifying meaningful deviations, and responding to emerging threats without relying on known attack signatures. As autonomous AI agents introduce new and unpredictable behaviors, that foundation becomes increasingly important to securing the enterprise.

Read the full report from the UK AI Security Institute here.

Continue reading
About the author
Adam Stevens
Senior Director of Product | Darktrace
あなたのデータ × DarktraceのAI
唯一無二のDarktrace AIで、ネットワークセキュリティを次の次元へ