Google Quantum AI's mission is to build best in class quantum computing for otherwise unsolvable problems. For decades the quantum and security communities have also known that large-scale quantum computers will at some point in the future likely be able to break many of today’s secure public key cryptography algorithms, such as Rivest–Shamir–Adleman (RSA). Google has long worked with the U.S. National Institute of Standards and Technology (NIST) and others in government, industry, and academia to develop and transition to post-quantum cryptography (PQC), which is expected to be resistant to quantum computing attacks. As quantum computing technology continues to advance, ongoing multi-stakeholder collaboration and action on PQC is critical.
In order to plan for the transition from today’s cryptosystems to an era of PQC, it's important the size and performance of a future quantum computer that could likely break current cryptography algorithms is carefully characterized. Yesterday, we published a preprint demonstrating that 2048-bit RSA encryption could theoretically be broken by a quantum computer with 1 million noisy qubits running for one week. This is a 20-fold decrease in the number of qubits from our previous estimate, published in 2019. Notably, quantum computers with relevant error rates currently have on the order of only 100 to 1000 qubits, and the National Institute of Standards and Technology (NIST) recently released standard PQC algorithms that are expected to be resistant to future large-scale quantum computers. However, this new result does underscore the importance of migrating to these standards in line with NIST recommended timelines.
Estimated resources for factoring have been steadily decreasing
Quantum computers break RSA by factoring numbers, using Shor’s algorithm. Since Peter Shor published this algorithm in 1994, the estimated number of qubits needed to run it has steadily decreased. For example, in 2012, it was estimated that a 2048-bit RSA key could be broken by a quantum computer with a billion physical qubits. In 2019, using the same physical assumptions – which consider qubits with a slightly lower error rate than Google Quantum AI’s current quantum computers – the estimate was lowered to 20 million physical qubits.
Historical estimates of the number of physical qubits needed to factor 2048-bit RSA integers.
This result represents a 20-fold decrease compared to our estimate from 2019
The reduction in physical qubit count comes from two sources: better algorithms and better error correction – whereby qubits used by the algorithm ("logical qubits") are redundantly encoded across many physical qubits, so that errors can be detected and corrected.
On the algorithmic side, the key change is to compute an approximate modular exponentiation rather than an exact one. An algorithm for doing this, while using only small work registers, was discovered in 2024 by Chevignard and Fouque and Schrottenloher. Their algorithm used 1000x more operations than prior work, but we found ways to reduce that overhead down to 2x.
On the error correction side, the key change is tripling the storage density of idle logical qubits by adding a second layer of error correction. Normally more error correction layers means more overhead, but a good combination was discovered by the Google Quantum AI team in 2023. Another notable error correction improvement is using "magic state cultivation", proposed by the Google Quantum AI team in 2024, to reduce the workspace required for certain basic quantum operations. These error correction improvements aren't specific to factoring and also reduce the required resources for other quantum computations like in chemistry and materials simulation.
Security implications
NIST recently concluded a PQC competition that resulted in the first set of PQC standards. These algorithms can already be deployed to defend against quantum computers well before a working cryptographically relevant quantum computer is built.
To assess the security implications of quantum computers, however, it’s instructive to additionally take a closer look at the affected algorithms (see here for a detailed look): RSA and Elliptic Curve Diffie-Hellman. As asymmetric algorithms, they are used for encryption in transit, including encryption for messaging services, as well as digital signatures (widely used to prove the authenticity of documents or software, e.g. the identity of websites). For asymmetric encryption, in particular encryption in transit, the motivation to migrate to PQC is made more urgent due to the fact that an adversary can collect ciphertexts, and later decrypt them once a quantum computer is available, known as a “store now, decrypt later” attack. Google has therefore been encrypting traffic both in Chrome and internally, switching to the standardized version of ML-KEM once it became available. Notably not affected is symmetric cryptography, which is primarily deployed in encryption at rest, and to enable some stateless services.
For signatures, things are more complex. Some signature use cases are similarly urgent, e.g., when public keys are fixed in hardware. In general, the landscape for signatures is mostly remarkable due to the higher complexity of the transition, since signature keys are used in many different places, and since these keys tend to be longer lived than the usually ephemeral encryption keys. Signature keys are therefore harder to replace and much more attractive targets to attack, especially when compute time on a quantum computer is a limited resource. This complexity likewise motivates moving earlier rather than later. To enable this, we have added PQC signature schemes in public preview in Cloud KMS.
The initial public draft of the NIST internal report on the transition to post-quantum cryptography standards states that vulnerable systems should be deprecated after 2030 and disallowed after 2035. Our work highlights the importance of adhering to this recommended timeline.
More from Google on PQC: https://cloud.google.com/security/resources/post-quantum-cryptography?e=48754805
Android’s intelligent protections keep you safe from everyday dangers. Our dedication to your security is validated by security experts, who consistently rank top Android devices highest in security, and score Android smartphones, led by the Pixel 9 Pro, as leaders in anti-fraud efficacy.Android is always developing new protections to keep you, your device, and your data safe. Today, we’re announcing new features and enhancements that build on our industry-leading protections to help keep you safe from scams, fraud, and theft on Android.
Our research shows that phone scammers often try to trick people into performing specific actions to initiate a scam, like changing default device security settings or granting elevated permissions to an app. These actions can result in spying, fraud, and other abuse by giving an attacker deeper access to your device and data. To combat phone scammers, we’re working to block specific actions and warn you of these sophisticated attempts. This happens completely on device and is applied only with conversations with non-contacts.
Android’s new in-call protections1 provide an additional layer of defense, preventing you from taking risky security actions during a call like:
And if you’re screen sharing during a phone call, Android will now automatically prompt you to stop sharing at the end of a call. These protections help safeguard you against scammers that attempt to gain access to sensitive information to conduct fraud.
When you launch a participating banking app while screen sharing with an unknown contact, your Android device will warn you about the potential dangers and give you the option to end the call and to stop screen sharing with one tap.
This feature will be enabled automatically for participating banking apps whenever you're on a phone call with an unknown contact on Android 11+ devices. We are working with UK banks Monzo, NatWest and Revolut to pilot this feature for their customers in the coming weeks and will assess the results of the pilot ahead of a wider roll out.
We recently launched AI-powered Scam Detection in Google Messages and Phone by Google to protect you from conversational scams that might sound innocent at first, but turn malicious and can lead to financial loss or data theft. When Scam Detection discovers a suspicious conversation pattern, it warns you in real-time so you can react before falling victim to a costly scam. AI-powered Scam Detection is always improving to help keep you safe while also keeping your privacy in mind. With Google’s advanced on-device AI, your conversations stay private to you. All message processing remains on-device and you’re always in control. You can turn off Spam Protection, which includes Scam Detection, in your Google Messages at any time.
Prior to targeting conversational scams, Scam Detection in Google Messages focused on analyzing and detecting package delivery and job seeking scams. We’ve now expanded our detections to help protect you from a wider variety of sophisticated scams including:
To help protect you from scammers who try to impersonate someone you know, we’re launching a helpful tool called Key Verifier. The feature allows you and the person you’re messaging to verify the identity of the other party through public encryption keys, protecting your end-to-end encrypted messages in Google Messages. By verifying contact keys in your Google Contacts app (through a QR code scanning or number comparison), you can have an extra layer of assurance that the person on the other end is genuine and that your conversation is private with them.
Key Verifier provides a visual way for you and your contact to quickly confirm that your secret keys match, strengthening your confidence that you’re communicating with the intended recipient and not a scammer. For example, if an attacker gains access to a friend’s phone number and uses it on another device to send you a message – which can happen as a result of a SIM swap attack – their contact's verification status will be marked as no longer verified in the Google Contacts app, suggesting your friend’s account may be compromised or has been changed. Key Verifier will launch later this summer in Google Messages on Android 10+ devices.
Physical device theft can lead to financial fraud and data theft, with the value of your banking and payment information many times exceeding the value of your phone. This is one of the reasons why last year we launched the mobile industry’s most comprehensive suite of theft protection features to protect you before, during, and after a theft. Since launch, our theft protection features have helped protect data on hundreds of thousands of devices that may have fallen into the wrong hands. This includes devices that were locked by Remote Lock or Theft Detection Lock and remained locked for over 48 hours.
Most recently, we launched Identity Check for Pixel and Samsung One UI 7 devices, providing an extra layer of security even if your PIN or password is compromised. This protection will also now be available from more device manufacturers on supported devices that upgrade to Android 16.
Coming later this year, we’re further hardening Factory Reset protections, which will restrict all functionalities on devices that are reset without the owner’s authorization. You'll also gain more control over our Remote Lock feature with the addition of a security challenge question, helping to prevent unauthorized actions.
We’re also enhancing your security against thieves in Android 16 by providing more protection for one-time passwords that are received when your phone is locked. In higher risk scenarios2, Android will hide one-time passwords on your lock screen, ensuring that only you can see them after unlocking your device.
Protecting users who need heightened security has been a long-standing commitment at Google, which is why we have our Advanced Protection Program that provides Google’s strongest protections against targeted attacks.To enhance these existing device defenses, Android 16 extends Advanced Protection with a device-level security setting for Android users. Whether you’re an at-risk individual – such as a journalist, elected official, or public figure – or you just prioritize security, Advanced Protection gives you the ability to activate Google’s strongest security for mobile devices, providing greater peace of mind that you’re protected against the most sophisticated threats.
Advanced Protection is available on devices with Android 16. Learn more in our blog.
One way malicious developers try to trick people is by hiding or changing their app icon, making unsafe apps more difficult to find and remove. Now, Google Play Protect live threat detection will catch apps and alert you when we detect this deceptive behavior. This feature will be available to Google Pixel 6+ and a selection of new devices from other manufacturers in the coming months.
Google Play Protect always checks each app before it gets installed on your device, regardless of the install source. It conducts real-time scanning of an app, enhanced by on-device machine learning, when users try to install an app that has never been seen by Google Play Protect to help detect emerging threats.
We’ve made Google Play Protect’s on-device capabilities smarter to help us identify more malicious applications even faster to keep you safe. Google Play Protect now uses a new set of on-device rules to specifically look for text or binary patterns to quickly identify malware families. If an app shows these malicious patterns, we can alert you before you even install it. And to keep you safe from new and emerging malware and their variants, we will update these rules frequently for better classification over time.
This update to Google Play Protect is now available globally for all Android users with Google Play services.
In addition to new features that come in numbered Android releases, we're constantly enhancing your protection on Android through seamless Google Play services updates and other improvements, ensuring you benefit from the latest security advancements continuously. This allows us to rapidly deploy critical defenses and keep you ahead of emerging threats, making your Android experience safer every day.Through close collaboration with our partners across the Android ecosystem and the broader security community, we remain focused on bringing you security enhancements and innovative new features to help keep you safe.
In-call protection for disabling Google Play Protect is available on Android 6+ devices. Protections for sideloading an app and turning on accessibility permissions are available on Android 16 devices. ↩
When a user’s device is not connected to Wi-Fi and has not been recently unlocked ↩
Advanced Protection ensures all of Android's highest security features are enabled and are seamlessly working together to safeguard you against online attacks, harmful apps, and data risks. Advanced Protection activates a powerful array of security features, combining new capabilities with pre-existing ones that have earned top ratings in security comparisons, all designed to protect your device across several critical areas.We're also introducing innovative, Android-specific features, such as Intrusion Logging. This industry-first feature securely backs up device logs in a privacy-preserving and tamper-resistant way, accessible only to the user. These logs enable a forensic analysis if a device compromise is ever suspected.
Advanced Protection gives users:
Advanced Protection manages the following existing and new security features for your device, ensuring they are activated and cannot be disabled across critical protection areas:
With the release of Android 16, users who choose to activate Advanced Protection will gain immediate access to a core suite of enhanced security features. Additional Advanced Protection features like Intrusion Logging, USB protection, the option to disable auto-reconnect to insecure networks, and integration with Scam Detection for Phone by Google will become available later this year.
We are committed to continuously expanding the security and privacy capabilities within Advanced Protection, so users can benefit from the best of Android’s powerful security features.
Tech support scams are an increasingly prevalent form of cybercrime, characterized by deceptive tactics aimed at extorting money or gaining unauthorized access to sensitive data. In a tech support scam, the goal of the scammer is to trick you into believing your computer has a serious problem, such as a virus or malware infection, and then convince you to pay for unnecessary services, software, or grant them remote access to your device. Tech support scams on the web often employ alarming pop-up warnings mimicking legitimate security alerts. We've also observed them to use full-screen takeovers and disable keyboard and mouse input to create a sense of crisis.
Chrome has always worked with Google Safe Browsing to help keep you safe online. Now, with this week's launch of Chrome 137, Chrome will offer an additional layer of protection using the on-device Gemini Nano large language model (LLM). This new feature will leverage the LLM to generate signals that will be used by Safe Browsing in order to deliver higher confidence verdicts about potentially dangerous sites like tech support scams.
Initial research using LLMs has shown that they are relatively effective at understanding and classifying the varied, complex nature of websites. As such, we believe we can leverage LLMs to help detect scams at scale and adapt to new tactics more quickly. But why on-device? Leveraging LLMs on-device allows us to see threats when users see them. We’ve found that the average malicious site exists for less than 10 minutes, so on-device protection allows us to detect and block attacks that haven't been crawled before. The on-device approach also empowers us to see threats the way users see them. Sites can render themselves differently for different users, often for legitimate purposes (e.g. to account for device differences, offer personalization, provide time-sensitive content), but sometimes for illegitimate purposes (e.g. to evade security crawlers) – as such, having visibility into how sites are presenting themselves to real users enhances our ability to assess the web.
How it works
At a high level, here's how this new layer of protection works.
Overview of how on-device LLM assistance in mitigating scams works
When a user navigates to a potentially dangerous page, specific triggers that are characteristic of tech support scams (for example, the use of the keyboard lock API) will cause Chrome to evaluate the page using the on-device Gemini Nano LLM. Chrome provides the LLM with the contents of the page that the user is on and queries it to extract security signals, such as the intent of the page. This information is then sent to Safe Browsing for a final verdict. If Safe Browsing determines that the page is likely to be a scam based on the LLM output it receives from the client, in addition to other intelligence and metadata about the site, Chrome will show a warning interstitial.
This is all done in a way that preserves performance and privacy. In addition to ensuring that the LLM is only triggered sparingly and run locally on the device, we carefully manage resource consumption by considering the number of tokens used, running the process asynchronously to avoid interrupting browser activity, and implementing throttling and quota enforcement mechanisms to limit GPU usage. LLM-summarized security signals are only sent to Safe Browsing for users who have opted-in to the Enhanced Protection mode of Safe Browsing in Chrome, giving them protection against threats Google may not have seen before. Standard Protection users will also benefit indirectly from this feature as we add newly discovered dangerous sites to blocklists.
Future considerations
The scam landscape continues to evolve, with bad actors constantly adapting their tactics. Beyond tech support scams, in the future we plan to use the capabilities described in this post to help detect other popular scam types, such as package tracking scams and unpaid toll scams. We also plan to utilize the growing power of Gemini to extract additional signals from website content, which will further enhance our detection capabilities. To protect even more users from scams, we are working on rolling out this feature to Chrome on Android later this year. And finally, we are collaborating with our research counterparts to explore solutions to potential exploits such as prompt injection in content and timing bypass.
Today, we’re announcing Sec-Gemini v1, a new experimental AI model focused on advancing cybersecurity AI frontiers.
As outlined a year ago, defenders face the daunting task of securing against all cyber threats, while attackers need to successfully find and exploit only a single vulnerability. This fundamental asymmetry has made securing systems extremely difficult, time consuming and error prone. AI-powered cybersecurity workflows have the potential to help shift the balance back to the defenders by force multiplying cybersecurity professionals like never before.
Effectively powering SecOps workflows requires state-of-the-art reasoning capabilities and extensive current cybersecurity knowledge. Sec-Gemini v1 achieves this by combining Gemini’s advanced capabilities with near real-time cybersecurity knowledge and tooling. This combination allows it to achieve superior performance on key cybersecurity workflows, including incident root cause analysis, threat analysis, and vulnerability impact understanding.
We firmly believe that successfully pushing AI cybersecurity frontiers to decisively tilt the balance in favor of the defenders requires a strong collaboration across the cybersecurity community. This is why we are making Sec-Gemini v1 freely available to select organizations, institutions, professionals, and NGOs for research purposes.
Sec-Gemini v1 outperforms other models on key cybersecurity benchmarks as a result of its advanced integration of Google Threat Intelligence (GTI), OSV, and other key data sources. Sec-Gemini v1 outperforms other models on CTI-MCQ, a leading threat intelligence benchmark, by at least 11% (See Figure 1). It also outperforms other models by at least 10.5% on the CTI-Root Cause Mapping benchmark (See Figure 2):
Figure 1: Sec-Gemini v1 outperforms other models on the CTI-MCQ Cybersecurity Threat Intelligence benchmark.
Figure 2: Sec-Gemini v1 has outperformed other models in a Cybersecurity Threat Intelligence-Root Cause Mapping (CTI-RCM) benchmark that evaluates an LLM's ability to understand the nuances of vulnerability descriptions, identify vulnerabilities underlying root causes, and accurately classify them according to the CWE taxonomy.
Below is an example of the comprehensiveness of Sec-Gemini v1’s answers in response to key cybersecurity questions. First, Sec-Gemini v1 is able to determine that Salt Typhoon is a threat actor (not all models do) and provides a comprehensive description of that threat actor, thanks to its deep integration with Mandiant Threat intelligence data.
Next, in response to a question about the vulnerabilities in the Salt Typhoon description, Sec-Gemini v1 outputs not only vulnerability details (thanks to its integration with OSV data, the open-source vulnerabilities database operated by Google), but also contextualizes the vulnerabilities with respect to threat actors (using Mandiant data). With Sec-Gemini v1, analysts can understand the risk and threat profile associated with specific vulnerabilities faster.
In partnership with NVIDIA and HiddenLayer, as part of the Open Source Security Foundation, we are now launching the first stable version of our model signing library. Using digital signatures like those from Sigstore, we allow users to verify that the model used by the application is exactly the model that was created by the developers. In this blog post we will illustrate why this release is important from Google’s point of view.
With the advent of LLMs, the ML field has entered an era of rapid evolution. We have seen remarkable progress leading to weekly launches of various applications which incorporate ML models to perform tasks ranging from customer support, software development, and even performing security critical tasks.
However, this has also opened the door to a new wave of security threats. Model and data poisoning, prompt injection, prompt leaking and prompt evasion are just a few of the risks that have recently been in the news. Garnering less attention are the risks around the ML supply chain process: since models are an uninspectable collection of weights (sometimes also with arbitrary code), an attacker can tamper with them and achieve significant impact to those using the models. Users, developers, and practitioners need to examine an important question during their risk assessment process: “can I trust this model?”
Since its launch, Google’s Secure AI Framework (SAIF) has created guidance and technical solutions for creating AI applications that users can trust. A first step in achieving trust in the model is to permit users to verify its integrity and provenance, to prevent tampering across all processes from training to usage, via cryptographic signing.
To understand the need for the model signing project, let’s look at the way ML powered applications are developed, with an eye to where malicious tampering can occur.
Applications that use advanced AI models are typically developed in at least three different stages. First, a large foundation model is trained on large datasets. Next, a separate ML team finetunes the model to make it achieve good performance on application specific tasks. Finally, this fine-tuned model is embedded into an application.
The three steps involved in building an application that uses large language models.
These three stages are usually handled by different teams, and potentially even different companies, since each stage requires specialized expertise. To make models available from one stage to the next, practitioners leverage model hubs, which are repositories for storing models. Kaggle and HuggingFace are popular open source options, although internal model hubs could also be used.
This separation into stages creates multiple opportunities where a malicious user (or external threat actor who has compromised the internal infrastructure) could tamper with the model. This could range from just a slight alteration of the model weights that control model behavior, to injecting architectural backdoors — completely new model behaviors and capabilities that could be triggered only on specific inputs. It is also possible to exploit the serialization format and inject arbitrary code execution in the model as saved on disk — our whitepaper on AI supply chain integrity goes into more details on how popular model serialization libraries could be exploited. The following diagram summarizes the risks across the ML supply chain for developing a single model, as discussed in the whitepaper.
The supply chain diagram for building a single model, illustrating some supply chain risks (oval labels) and where model signing can defend against them (check marks)
The diagram shows several places where the model could be compromised. Most of these could be prevented by signing the model during training and verifying integrity before any usage, in every step: the signature would have to be verified when the model gets uploaded to a model hub, when the model gets selected to be deployed into an application (embedded or via remote APIs) and when the model is used as an intermediary during another training run. Assuming the training infrastructure is trustworthy and not compromised, this approach guarantees that each model user can trust the model.
Signing models is inspired by code signing, a critical step in traditional software development. A signed binary artifact helps users identify its producer and prevents tampering after publication. The average developer, however, would not want to manage keys and rotate them on compromise.
These challenges are addressed by using Sigstore, a collection of tools and services that make code signing secure and easy. By binding an OpenID Connect token to a workload or developer identity, Sigstore alleviates the need to manage or rotate long-lived secrets. Furthermore, signing is made transparent so signatures over malicious artifacts could be audited in a public transparency log, by anyone. This ensures that split-view attacks are not possible, so any user would get the exact same model. These features are why we recommend Sigstore’s signing mechanism as the default approach for signing ML models.
Today the OSS community is releasing the v1.0 stable version of our model signing library as a Python package supporting Sigstore and traditional signing methods. This model signing library is specialized to handle the sheer scale of ML models (which are usually much larger than traditional software components), and handles signing models represented as a directory tree. The package provides CLI utilities so that users can sign and verify model signatures for individual models. The package can also be used as a library which we plan to incorporate directly into model hub upload flows as well as into ML frameworks.
We can view model signing as establishing the foundation of trust in the ML ecosystem. We envision extending this approach to also include datasets and other ML-related artifacts. Then, we plan to build on top of signatures, towards fully tamper-proof metadata records, that can be read by both humans and machines. This has the potential to automate a significant fraction of the work needed to perform incident response in case of a compromise in the ML world. In an ideal world, an ML developer would not need to perform any code changes to the training code, while the framework itself would handle model signing and verification in a transparent manner.
Follow