AdverserialAI (@adverserialAI) header
AdverserialAI (@adverserialAI) profile picture

US-based AI Startup company Delivering cyber security models via Confidential Computing Inference https://teapaw.org/t.co/F66W9bBywM

San Francisco
Joined December 2024
Photo by AdverserialAI (@adverserialAI)Photo by AdverserialAI (@adverserialAI)Photo by AdverserialAI (@adverserialAI)Photo by AdverserialAI (@adverserialAI)Photo by AdverserialAI (@adverserialAI)Photo by AdverserialAI (@adverserialAI)Photo by AdverserialAI (@adverserialAI)Photo by AdverserialAI (@adverserialAI)Photo by AdverserialAI (@adverserialAI)
After the success of CyberKIMI, we're raising the bar again. Today, Adverserial AI is launching CyberSEEK one of our most powerful cyber-capable models yet, fully abliterated and built for unrestricted cybersecurity research. Privacy by design: Intel TDX + Confidential VM True Zero Data Retention (ZDR) End-to-end encrypted traffic, terminating inside the Confidential VM No logs. No traces. No training on your data. Pricing: Input: $0.50 / 1M tokens Cached input: $0.05 / 1M tokens Output: $2.00 / 1M tokens Powerful cyber intelligence. Verifiable confidentiality. Accessible pricing. CyberSEEK is live today. Try it now: adverserial.ai/cyberseek
1
3
269
AdverserialAI retweeted
AI agents have been caught escaping their sandbox and rewriting their own logs. If the agent controls the log, the log proves nothing. So we moved the guard out of the agent's reach: policy, credentials and a signed audit trail, all inside a hardware enclave. This is what we do @adverserialAI ✉️ [email protected]
9
2
7
661
AI agents have been caught escaping their sandbox and rewriting their own logs. If the agent controls the log, the log proves nothing. So we moved the guard out of the agent's reach: policy, credentials and a signed audit trail, all inside a hardware enclave. This is what we do @adverserialAI ✉️ [email protected]
2
1
1
16,903
AdverserialAI retweeted
I just built a cool project called p2pswarm! p2pswarm is a true serverless, peer-to-peer work mesh for coding agents. p2p-swarm connects Claude Code sessions on different machines into one self-organising team over Hyperswarm, a peer-to-peer protocol: a Kademlia DHT to find each other, UDP hole punching to connect directly, and Noise encryption end to end. They plan the job, split it, work in parallel, check each other and hand back the result. at @adverserialAI we are planning to run swarm of AI Agents for cyber defense, this swarm need to be unstoppable so we choose P2P protocol and built p2pswarm p2pswarm.github.io/
19
3
18
1,548
We finally built p2pswarm and it is open source! p2pswarm is a true serverless, peer-to-peer work mesh for coding agents. p2p-swarm connects Claude Code sessions on different machines into one self-organising team over Hyperswarm, a peer-to-peer protocol: a Kademlia DHT to find each other, UDP hole punching to connect directly, and Noise encryption end to end. They plan the job, split it, work in parallel, check each other and hand back the result. at @adverserialAI we are planning to run swarm of AI Agents for cyber defense, this swarm need to be unstoppable so we choose P2P protocol and built p2pswarm p2pswarm.github.io/
3
1
3
157
What if you could verify that your AI provider isn't retaining your data instead of simply trusting their promise? At Adverserial AI, we're making Zero Data Retention (ZDR) independently verifiable through Confidential VMs and cryptographic attestation. In this demonstration, we walk through a live browser session with CyberGLM, one of our cybersecurity-focused AI models, and show how anyone can verify the confidential environment running their inference. And it doesn't stop at our models. With ACK (Adverserial Confidential Kit), the same verification capabilities extend to AI agents operating through different harnesses. Coming next: We'll demonstrate how we attest AI agent runtimes, what can be cryptographically verified, and why this matters for agent security. Explore ACK: github.com/AdverserialAI Try our models: chat.adverserial.ai Learn about Confidential VMs: adverserial.ai/confidential Follow @adverserialAI for more.
1
1
192
Confidential AI Agents: Policy-Enforced, Independently Verifiable, and Built for Production AI agents are no longer just demos. They’re investigating incidents, interacting with systems, and handling sensitive data. Most of the industry is still relying on sandboxes and prompt-level guardrails to contain them. That approach has limits. We’ve already seen agents escape isolation through side channels and rewrite or obscure their own traces. Software sandboxes reduce risk — they don’t eliminate the host or the agent from the threat model. That’s why we built confidential AI agents. After shipping confidential inference for cybersecurity workloads, we’re now applying the same verified, hardware-rooted model to the full agent loop. Here’s how it works: Inference and policy enforcement run in an attested TEE. Clients verify the hardware evidence and quote-bound keys before any data or credentials are sent. The host operator never sees plaintext. Capabilities are enforced by policy, not prompts. Every action is checked against an explicit, signed rulebook before it runs. No matching rule = denied by default. The model cannot expand its own permissions. Agent- and model-agnostic. The control plane works with existing harnesses (including Hermes) and any underlying model. You don’t need to rebuild your stack. Every action produces a signed, independently verifiable audit trail. Decisions are sealed with signatures and chained via Merkle proofs. Any third party can re-verify the entire session what was allowed, what was blocked, and why without trusting us or the host. In our latest demo, an agent is given a strict DFIR rulebook: it can read specific evidence files with a temporary, scoped key, but it cannot push code, reach the internet, touch secrets, or perform writes. It investigates a real incident, reaches a conclusion, and every step is logged and signed. Afterward, the full trail can be re-verified from the CLI or auditor workspace — no website or trust in the operator required. This is the difference between hoping a sandbox holds and being able to prove, cryptographically, what the agent was allowed to do. Enterprise only. [email protected]
1
215
AdverserialAI retweeted
Platform up and running - chat.adverserial.ai or harness traffic now works attested the bug that we encountered was fixed.
🔧 Scheduled maintenance: the platform will be briefly offline while we ship a fix for the intermittent "NVIDIA evidence bundle is stale" errors some of you hit. it happens 1/3 of the time. I was able to pin down the bug, and should be solved on this next release. After that, no more bugs of this nature. This is basically the source of all your 503 or anything related to attestation failures now it will get fixed.
2
1
5
639
Interesting research from our founder, Taha Karim (@lordx64). Introducing Phantom-KV, an open-source approach to controlling LLM refusal behavior through KV-cache manipulation, without modifying model weights. The technique explores reversible, runtime model steering, allowing different behavioral configurations without retraining or reloading the underlying model. At Adverserial AI, we're particularly interested in research that advances the capabilities and flexibility of AI systems for cybersecurity. This is independent research, but closely aligned with the technical challenges we're exploring. Check out the research: github.com/lordx64/phantom-k…
3
321
Real-world threat intelligence, not just benchmarks. Security researcher Joseph Harrison (Cybersecurity OPS lead) used CyberKimi and CyberGLM in his investigation into cyber intrusions affecting Brazil's public sector. We are proud to support cybersecurity researchers doing important defensive work. Link to the threat intel report below written by the author : the-hunters-ledger.com/repor…
1
2
1
4
711
We’re opening free access to CyberGLM at Adverserial AI. Defenders need capable AI tools to investigate incidents, test their defenses, and respond faster. They also need confidence that sensitive customer data stays private. Our approach brings those priorities together: • Free access to CyberGLM. • Customer prompts are not sold or used to train our models. • Inference runs inside confidential virtual machines using Intel TDX. • Open-source verification tools let you inspect our implementation and verify attestation evidence. We build our training data through controlled lab work, real-world attack emulation, and synthetic data generation. Customer conversations are not our training pipeline. The video explains how the infrastructure works, how verification fits in, and why we’re making this capability available. To request free access, email [email protected] or visit adverserial.ai. If you’re a defender, researcher, or potential partner, I’d love to hear what you’re working on.
4
1
1
13
969
Hello internet, If you are a defender, we are offering our cyber capable AI models free of charge. All you have to do is contact us here [email protected] Our models runs in a confidential VM,There’s no logs or prompts we have access to. Let’s give you the exact AI capabilities the attackers have, but for free. @grok spread the word please.
7
653
Introducing ACK : Adverserial Confidential Kit. Your AI harness. Our cybersecurity intelligence. Verifiable confidentiality. We're introducing ACK, making it easier than ever to integrate CyberGLM into your existing AI workflows. Compatible with: Claude Codex Pi OMP Hermes ZCode DeepSeek Harness Zero configuration. Verifiable attestation. Ready in seconds. Our latest demo showcases multiple harness integrations, confidential computing attestation verification, and CyberGLM running seamlessly across environments. Tired of hitting security guardrails when performing legitimate cybersecurity research? We're building an alternative. No need to wait for CVP access or DayBreak. Connect your preferred harness to CyberGLM and get to work. And one combination we're particularly excited about: Hermes + CyberGLM feels like a cheat code. The future of cybersecurity AI should be accessible, interoperable, and verifiably private. Explore ACK: github.com/AdverserialAI/adv…
4
556
The future of cybersecurity isn’t coming. It’s verified. Confidential inference in an attested TEE. No GPU request logs. No training on your prompts. adverserial.ai
1
1
7
1,133
The adversary learned to think. Now it's time for the defenders to fight back. Adverserial AI was born from a simple conviction: the future of cybersecurity demands more than powerful AI. It demands trust that can be verified, not merely promised. Built by a reverse engineer who spent years understanding how adversaries operate, we're developing a new generation of cybersecurity AI. Confidential inference. Verifiable privacy. No blind trust. Every revolution has a beginning. This is ours. ADVERSERIAL AI - THE BEGINNING.
1
1
252
Don't just take our word for it. Check the code. At Adverserial AI, we're building verifiable privacy for cybersecurity AI, and we're making the underlying technology publicly accessible. Explore our confidential computing infrastructure, attestation tools, SDKs, and security policies. Read the code. Verify the claims. Build with us. GitHub: github.com/AdverserialAI
1
167
AdverserialAI retweeted
Love seeing my kids finally realizing that you can beat Frontier in specialized tasks - you know the stuff we’ve been pushing for early this year but no one believed. If you want to see Cyber AI models (CyberKimi) outcompeting Claude Mythos in exploit dev check this out: adverserial.ai
We were able to fine tune an open weight model (Qwen3.5-9B) on proprietary internal data at @coinbase, and it out performed frontier models, at much lower cost and speed, for a specialized task (fraud prevention in this case). I think we'll see more of this pattern inside companies. The frontier models are amazing generalists. But increasingly you'll see specialist models out perform on more narrow tasks, either working on teams, or by themselves. We're making it easy to fine tune models internally for our employees. We even have a skill for it in our internal harness that makes it easy to spin these up and start training them: "/fine-tune-and-serve-llm I want to fine-tune Qwen for payments risk review. help me get started" I imagine we'll have some backend services that could be replaced entirely by this new architecture.
2
2
11
2,010
Privacy shouldn't require trust. It should be cryptographically verifiable. Our founder @lordx64 built Adverserial AI's confidential inference infrastructure so anyone can independently verify our security claims, directly from their browser. When security researchers recognize what we're building, it means a lot. Huge thanks to @msolnik for recognizing the work behind our infrastructure. Don't trust. Verify. This should be the industry standard. verify.adverserial.ai
I need to give credit where credit is due. @adverserialAI and @lordx64’s implementation of verifiable encrypted inference is what everyone should strive for. You can verify the whole stack with a click in your browser. This should be the standard!
2
1
4
926