At AirAsia, we operate a fleet of more than 270 aircraft across 23 markets, fly to more than 150 destinations and carry 100m guests each year. We’ve also been named the world’s best low-cost carrier for 11 years running. To accomplish all of this, we rely heavily on our 22,000 Allstars (employees). As AirAsia co-founder Tony Fernandes likes to say, “it has always been about the people.”

With Cloud Identity in place, our IT department can spend less time worrying about managing multiple on-premises directory servers and and can instead focus on delivering value to our Allstar employees.
–Declan Hogan, Group CIO, AirAsia

In my role as CIO, it’s critical that I give our Allstars the tools and technology they need to get their jobs done, while at the same time ensuring that our company’s data is protected and secure. While this is challenging enough in normal circumstances, we’re also in the midst of rapidly moving from legacy on-premises technologies to the cloud. Google Cloud has been a critical partner for us in this journey.

Identity Challenges

AirAsia, like many other enterprises, has relied on a legacy on-premises directory for many years. As our company has quickly grown and expanded to new markets and regions, we’ve had to manage multiple servers across a number of on-premises data centers and the public cloud, which has proved costly and time-consuming.

Our Allstars, located all across Asia, need to easily access a number of legacy on-premises apps in addition to a growing number of SaaS apps. As a business, we also needed a more seamless integration between our HR system of record and our identity solution for user provisioning and lifecycle management.

Deploying Cloud Identity is a key step towards enabling the BeyondCorp (or zero trust) security model, which we feel is the best approach to strengthen our security posture and fight modern threats.
–Declan Hogan, Group CIO, AirAsia

Solving these challenges with our existing on-premises directory was simply not feasible for us.

In recent years, we partnered with Google Cloud to help drive our digital transformation, including moving dozens of workloads and apps to Google Cloud Platform (GCP), deploying G Suite as our collaboration and productivity solution for all of our employees, and replacing thousands of Windows laptops with fast and secure Chromebooks.

We brought up our identity concerns with the Google Cloud team, and after a number of conversations, we decided to deploy Cloud Identity, Google’s cloud-based identity and access management solution, to help address the identity challenges we were facing.

Why Cloud Identity for AirAsia?

We ended up choosing Cloud Identity for a few key reasons. Here at AirAsia, we are eager to move to the cloud as quickly as possible, and moving identity management to the cloud is a key enabler of this and our broader digital transformation. Managing identities from the cloud also enables us to have a single identity and set of credentials for each employee, which they can use to access all of the applications they need to be productive, both in the cloud and on-premises.

In addition, deploying Cloud Identity is a key step towards enabling the BeyondCorp (or zero trust) security model, which we feel is the best approach to strengthen our security posture and fight modern threats.

Cloud Identity also integrates seamlessly with our existing technologies, which includes not only Google Cloud products like GCP, G Suite, and Chrome OS, but also third party tools like Citrix, Papercut, and others.

And finally, Cloud Identity offered us significant cost and resource savings. With Cloud Identity in place, our IT department can spend less time worrying about managing multiple on-premises directory servers and and can instead focus on delivering value to our Allstar employees.

As the way people work continues to evolve, keeping security policies in place that protect organizations but give workers the ability to get things done is more important than ever. IT and security teams must aim to stay a step ahead of web-based security threats that come their organization’s way. To help, the Center for Internet Security (CIS) team has released the latest CIS Benchmark 2.1 for Google Chrome. This Benchmark offers independent recommendations on which Chrome policies to configure to help support organizations’ security and compliance needs. Thanks to Chrome being built with security at its core, in many cases, Chrome default settings are aligned with CIS recommendations.

Chrome is secure by default, but we also pride ourselves on providing customizations for enterprises to allow Chrome to better fit the needs of their business. And with hundreds of policies available through Chrome Browser Cloud Management and Group Policy Objects (Note: The CIS Benchmark is also available as a GPO), organizations can do just that. 

Throughout the CIS guide you’ll notice that there are different designations for configuration profiles. Any labeled Level 1, are considered to be a good baseline for an organization. Level 2 profiles are recommended for deployments that require the highest level of security, but note that these settings could have a trade off on user productivity. We recommend looking at each setting and determining if it’s a good fit for your business. 

The benchmark is made up of five sections:

• Enforced Defaults — Notes policies that are configured by default when you install Chrome. Enforcing these settings at an enterprise level can prevent these settings from being changed by business users to less secure options.
• Attack Surface Reduction — Details how to disable web features that may not be necessary in your enterprise environment and could reduce your overall attack surface.
• Privacy — Surfaces settings that improve user privacy.
• Data Loss Prevention — Contains settings that can help prevent data loss and protect your organization’s data. (Note: These recommendations cover additional capabilities that can be added to Chrome through BeyondCorp Enterprise). 
• Forensics (Post Incident) — Shares recommendations on policies that give insights into post incident forensics and analysis. 

Organizations can use these benchmarks to optimize the best way to secure Chrome in their environment. Download the CIS Benchmark here and check out our team’s configuration guide for additional recommendations on how to configure Chrome.

Note: This benchmark was created using a consensus review process composed of subject matter experts. Consensus participants provide perspective from a diverse set of backgrounds such as consulting, software development, audit and compliance, security  research, operations, government, and legal. While these recommendations come from a trusted source, it’s important for each organization to weigh which policies make the most sense for their business.

The future of AI is better AI—designed with ethics and responsibility built in from the start. This means putting the brakes on AI-driven transformation until you have a well-functioning strategy and process in place to ensure your models deliver fair outcomes. Failing to recognize this imperative is a threat to your bottom line. The following post provides a simple framework to follow to keep your business on the right track as you place more trust in algorithms. 

AI is inherently sociotechnical. AI systems represent the interconnectedness of humans and technology. They are designed to be used by and to inform humans within specific contexts, and the speed and scale of AI means that any lack of responsibility—such as bias, safety, privacy, scientific excellence etc—will also replicate at that same speed and scale. Without ethics and responsibility built in by design, AI systems lack the critical “inputs” or societal context that enable long term success. 

Lawsuits stemming from AI systems that are biased towards certain groups are stacking up. In August 2020, IBM was forced to settle a lawsuit with the city of Los Angeles for misappropriating data it collected for its weather channel app. Health services company, Optum, is being investigated by regulators for creating an algorithm that allegedly recommended that doctors and nurses pay more attention to white patients than to sicker black patients. And Facebook, which granted Cambridge Analytica, a political firm, access to the personal data of more than 50 million people, is buried in legal work.  Google has also run into its share of issues with algorithms making egregious mistakes. 

While lawsuits are real, the foundational reason ethical AI is critical to your bottom line is trust. Without it, increasingly, consumers will ignore you and choose a brand they do trust. Research from Kantar, which runs one of the largest global brand equity studies (4 million consumers, 18,000 brands, across 50 markets), revealed that almost 9% of a brand’s equity is driven by corporate reputation, of which responsibility is a key attribute. Over the last decade, the importance of responsibility to consumers in relation to making brand choices has tripled. 

The study stated brands perceived to be among the world’s most trusted and responsible shared three crucial factors that proved particularly important for building consumer trust and confidence, even when a brand might be new to a market. These are:

• Honesty and openness
• Respect and inclusion
• Identifying with and caring for customers

Brands that develop these associations more strongly tend to outperform their competitors in defending and growing their brand value.

Technology and business leaders need to focus on four areas to accomplish a well-functioning ethical AI strategy. Lopez Research refers to this group of tasks as SEED, which stands for security, ethics, explainability, and data (SEED). Each of these topics could be an article in itself, but this post will define several essential components. 

SECURITY (S)  

It might not seem obvious, but a robust AI strategy requires an embedded security strategy. Companies should look for hardware-level security in components such as GPUs and CPUs. IT leaders should build software security into models to minimize attacks such as poisoning, evasion, deepfakes, backdoors, and model extraction. The threat of adversarial data poisoning attacks machine learning models by maliciously introducing inaccurate data designed to corrupt the model’s ability to be accurate. Another security threat is model extraction, also known as model cloning, where a hacker finds a way to either reconstruct a black-box machine learning model or extract the training data. The first line of defense against all security attacks is to design security at the outset, but the next best step is to frequently test models to ensure they are operating as planned. Business leaders, data science experts, and IT leaders must work together to regularly review the outcomes of AI models.

ETHICS (E)

Today, organizations must understand that ethics should be designed into the solution at its outset. The ethics process starts with defining the potential positive and negative outcomes of the model that your business is creating. Once the team has evaluated potential harmful effects, which means unpacking the systems, beliefs, power hierarchies and dynamics that interconnect with the technology, it’s your responsibility to eliminate or minimize the impact of these outcomes. It’s also critically important to review the impact of models in production and shut down models demonstrating issues. An example of this was the public beta release of the Tay chatbot that Microsoft deployed and rapidly shut down because it propagated negative biases. 

Yet, many organizations aren’t taking this action. The FICO study revealed that 93% of companies said responsible AI was critical for success but only 33% of these companies were measuring AI model outputs to ensure these models were operating as expected (measuring for model drift). Another survey by Pew Research revealed that 68% believe that ethical principles focused primarily on the public good will not be employed in most AI systems by 2030. 

Regulations may turn this tide, regardless of whether organizations plan to adopt an ethical AI framework. Laws governing the ethical use of data in AI are expected to be finalized as soon as 2022, such as the European Commission’s proposed legal framework for AI. Organizations that start with ethical use of AI in mind will be better positioned to deal with customer privacy concerns and regulatory compliance.

EXPLAINABILITY(E)

As models have become more sophisticated, it’s also become increasingly difficult to explain why a model created a specific outcome. In the FICO Responsible AI  report, 65% of respondents could not explain how specific AI model decisions or predictions are made, and only 35% said their organization made an effort to use AI in a way that was transparent and accountable.  However, it’s never been more important to clarify how AI models came to conclusions such as why a loan was denied, why a particular strategy should be implemented, and how AI selected a set of resumes to review for a position. The goal is to create an explainable AI model from the outset but many of today’s models lack this capability. Every business should review its existing models and use open-source toolkits that can be found on Github.com that support the interpretability and explainability of machine learning models. 

Keep in mind that explainability isn’t one-size-fits-all. Different stakeholders need different types of information. Much of explainability to date has focused on “opening the black box” which gets equated to information that is only useful for other data scientists. That’s important, but it doesn’t help the line of business users whose workflows AI is integrated into, or end users who deserve information about how decisions are made; or policymakers who don’t have data science backgrounds, and so on. 

DATA (D) 

An equally important item in ethics is data. Ethics starts with ensuring you have the correct data to create and update models. Three main issues include representative data, inherent biases within existing data, and inaccurate data. A critical issue that most companies miss in creating models is that current data sets frequently lack full market representation. A recent Capgemini Research Institute report revealed that 65% of executives “were aware of the issue of discriminatory bias” with these systems.

Awareness is the first step, but organizations must take action to remedy this issue. Historical data may no longer serve a company’s current needs for model creation. Historical records may contain biases against certain groups. For example, historical criminal data records show an imbalance in ethnic groups’ incarceration, which would lead to model biases. Additionally, laws and societal norms also change. Certain groups were prosecuted for sexual preference in the past, but today this information would create an inaccurate model. 

Companies have also discovered that using demographic data, a common practice in marketing, can also lead to model bias. For example, individuals that primarily used cash for transactions and others that lived in specific zip codes were at a disadvantage in banking models to determine creditworthiness. To minimize these issues, a company needs to augment its data with full representation in areas such as ethnicity, gender, age, behavioral and economic profiles.

Design AI models with a continuous feedback loop

Another, more prominent, yet tricky issue is data accuracy. As the adage says, garbage in, equals garbage out. The least appreciated but arguably the most essential component of the AI model lifecycle is ensuring the model has accurate data at all times. Inaccurate data from either poor data hygiene or data that was tampered with for security purposes can cause model failures. Organizations need to invest the time and resources to ensure they have the correct data. Data privacy is another key element that businesses must address, but the concepts of data privacy, sovereignty, and security are significant enough that we will come back to this in a separate article. 

Overall, it’s clear that while we may have an abundance of data, it most likely doesn’t represent what we want to model for the future. A successful AI strategy is an ethical AI strategy that requires the organization to be thoughtful in its model creation by ensuring it has a broad representation of accurate data and testing the outcomes to ensure the models are secure and operating as expected. 

Organizations that define an AI model lifecycle with a continuous feedback loop will reap the benefits of better intelligence. This will increasingly mean stronger, longer lasting trust with customers and staying on the right side of new laws and regulations.

The cybersecurity industry is faced with the tremendous challenge of analyzing growing volumes of security data in a dynamic threat landscape with evolving adversary behaviors. Today’s security data is heterogeneous, including logs and alerts, and often comes from more than one cloud platform. In order to better analyze that data, we’re excited to announce the release of the Cloud Analytics project by the MITRE Engenuity Center for Threat-Informed Defense, and sponsored by Google Cloud and several other industry collaborators.

Since 2021, Google Cloud has partnered with the Center to help level the playing field for everyone in the cybersecurity community by developing open-source security analytics. Earlier this year, we introduced Community Security Analytics (CSA) in collaboration with the Center to provide pre-built and customizable queries to help detect threats to your workloads and to audit your cloud usage. The Cloud Analytics project is designed to complement CSA.

The Cloud Analytics project includes a foundational set of detection analytics for key tactics, techniques and procedures (TTPs) implemented as vendor-agnostic Sigma rules, along with their adversary emulation plans implemented with CALDERA framework. Here’s a overview of Cloud Analytics project, how it complements Google Cloud’s CSA to benefit threat hunters, and how they both embrace Autonomic Security Operations principles like automation and toil reduction (adopted from SRE) in order to advance the state of threat detection development and continuous detection and response (CD/CR).

Both CSA and the Cloud Analytics project are community-driven security analytics resources. You can customize and extend the provided queries, but they take a more do-it-yourself approach—you’re expected to regularly evaluate and tune them to fit your own requirements in terms of threat detection sensitivity and accuracy. For managed threat detection and prevention, check out Security Command Center Premium’s realtime and continuously updated threat detection services including Event Threat Detection, Container Threat Detection, and Virtual Machine Threat Detection. Security Command Center Premium also provides managed misconfiguration and vulnerability detection with Security Health Analytics and Web Security Scanner.

Cloud Analytics vs Community Security Analytics

Similar to CSA, Cloud Analytics can help lower the barrier for threat hunters and detection engineers to create cloud-specific security analytics. Security analytics is complex because it requires:

• Deep knowledge of diverse security signals (logs, alerts) from different cloud providers along with their specific schemas;
• Familiarity with adversary behaviors in cloud environments;
• Ability to emulate such adversarial activity on cloud platforms;
• Achieving high accuracy in threat detection with low false positives, to avoid alert fatigue and overwhelming your SOC team.

The following table summarizes the key differences between Cloud Analytics and CSA:

Together, CSA and Cloud Analytics can help you maximize your coverage of the MITRE ATT&CK® framework, while giving you the choice of detection language and analytics engine to use. Given the mapping to TTPs, some of these rules by CSA and Cloud Analytics overlap. However, Cloud Analytics queries are implemented as Sigma rules which can be translated to vendor-specific queries such as Chronicle, Elasticsearch, or Splunk using Sigma CLI or third party-supported uncoder.io, which offers a user interface for query conversion. On the other hand, CSA queries are implemented as YARA-L rules (for Chronicle) and SQL queries (for BigQuery and now Log Analytics). The latter could be manually adapted to specific analytics engines due to the universal nature of SQL

Getting started with Cloud Analytics

To get started with the Cloud Analytics project, head over to the GitHub repo to view the latest set of Sigma rules, the associated adversary emulation plan to automatically trigger these rules, and a development blueprint on how to create new Sigma rules based on lessons learned from this project.

The following is a list of Google Cloud-specific Sigma rules (and their associated TTPs) provided in this initial release; use these as examples to author new ones covering more TTPs.

Sigma rule example

Using the canonical use case of detecting when a storage bucket is modified to be publicly accessible, here’s an example Sigma rule (copied below and redacted for brevity):

The rule specifies the log source (gcp.audit), the log criteria ( storage.googleapis.com service and storage.setIamPermissions method) and the keywords to look for (allUsers, ADD) signaling that a role was granted to all users over a given bucket. To learn more about Sigma syntax, refer to public Sigma docs.

However, there could still be false positives such as a Cloud Storage bucket made public for a legitimate reason like publishing static assets for a public website. To avoid alert fatigue and reduce toil on your SOC team, you could build more sophisticated detections based on multiple individual Sigma rules using Sigma Correlations.

Using our example, let’s refine the accuracy of this detection by correlating it with another pre-built Sigma rule which detects when a new user identity is added to a privileged group. Such privilege escalation likely occurred before the adversary gained permission to modify access of the Cloud Storage bucket. Cloud Analytics provides an example of such correlation Sigma rule chaining these two separate events.

What’s next

The Cloud Analytics project aims to make cloud-based threat detection development easier while also consolidating collective findings from real-world deployments. In order to scale the development of high-quality threat detections with minimum false positives, CSA and Cloud Analytics promote an agile development approach for building these analytics, where rules are expected to be continuously tuned and evaluated.

We look forward to wider industry collaboration and community contributions (from rules consumers, designers, builders, and testers) to refine existing rules and develop new ones, along with associated adversary emulations in order to raise the bar for minimum self-service security visibility and analytics for everyone.

Acknowledgements