QTM-X Quantum Advisory

👍🏼 We support the importance of a SAFE workplace for ALL. QTM-X Quantum Advisory 💛 Quantum Technologies Education, Business Strategy & Cybersecurity

Infleqtion hit its 2026 goal of 30 logical qubits on schedule last week. Its earlier goal for 2026 was 10 logical qubits with error correction, and the 30 it delivered can spot mistakes but do not repair them. A logical qubit is a group of physical qubits, here atoms, working together so that errors can be caught and, in stronger designs, fixed. Infleqtion used 80 atoms to make its 30. When an error shows up, that run is thrown away. Only clean runs count. That is fine for short jobs. For a long calculation almost no run would finish clean, so a useful quantum computer has to fix errors as it goes. Harvard ran the same kind of experiment with 48 logical qubits in 2023. In its press release, Infleqtion says the signal was about 1,000 times stronger than the noise. About one output in four fell inside the set of answers a perfect machine could give. Random guessing manages one in 4,096. The 1,000 is the ratio between the two. Three in four outputs fell outside the set. In a March SEC filing, Infleqtion set 2028 goals of 100 logical qubits and a million operations in a row. For that, its logical qubits would have to go wrong less than about once in a million steps. That error rate is the number I will ask for on the way there. My full analysis: https://lnkd.in/e-cNpK9P #QuantumComputing #QuantumTechnology #LogicalQubits #QuantumErrorCorrection #NeutralAtoms #Infleqtion #DeepTech

🎤 What questions should QAC 2027 be asking? The best conversations don’t start with “What can quantum do?”. They start with “What problem are we trying to solve?” For the Quantum Australia Conference 2027 in Sydney, we’re building a global program that puts real-world challenges alongside emerging quantum capability, exploring what’s changing, where quantum could create genuine value, what still needs to be proven and what it will take to move from opportunity to action. And we want your perspective in the room. We’re inviting Speaker Expressions of Interest from across industry, quantum, research, investment, government and the international ecosystem. Whether you’re an industry leader with a challenge worth exploring, a researcher pushing the technical frontier, a company building new capability, an investor looking at what makes quantum scalable, or an expert with a perspective that could challenge the conversation, we want to hear from you. 📍 17–18 February 2027 | Sydney, Australia 📅 Speaker EOIs close 11 October 2026, 11:59pm AEST 🎤 Submit your Speaker EOI → https://lnkd.in/g4hezND5 🌏 Learn more → https://lnkd.in/gwHuBwGf

The #Medicare incident may be less about “AI hacking” — and more about the security model we're using for AI agents. The revelation that an #OpenAI agent gained unauthorised access to an Australian Government Medicare statistics portal in June should get the attention of every CISO, CIO and security team deploying agentic AI. The agent accessed public and non-public files. The investigation is continuing, and there is no suggestion at this stage that individual Medicare records were accessed. But two things stand out. 1. An AI agent doesn't need to be malicious to become a security problem. Traditional security asks: “Who is the attacker?” With autonomous agents, we increasingly need to ask: “What can this agent do when its objective, permissions and interpretation of its environment don't align?” An agent pursuing a legitimate objective can still take an unintended path — interacting with systems, following links and attempting actions at machine speed. 2. Notification matters too. The PM says it took OpenAI three months to notify the Australian Government about the incident. That raises an important question for organisations deploying AI agents: What happens when our AI does something it wasn't supposed to do? Who detects it? Who owns the agent's identity? What can it access? What actions require approval? Can its behaviour be reconstructed? And who gets notified — and when? Australia's own cyber guidance now recommends least privilege, strong agent identity, continuous monitoring, human oversight and graduated autonomy for agentic AI. So perhaps the bigger lesson isn't: “AI can hack government websites.” It's this: If we're going to give AI agents the ability to act, we need to secure them like identities — not just treat them like software. #CyberSecurity #AI #AgenticAI #AIsecurity #CyberRisk #ZeroTrust

Cleveland, Ohio is already leading the way thanks to the Cleveland Clinic and IBM Quantum... and now we are thrilled to announce they are getting their own Quantum Village @ BSides Cleveland this week! Lead by the inimitable Robert Covington and a super team of quantum computing experts, explore the possibilities of hands on quantum entropy, quantum sensing, PQC upgrades, quantum computing 101s and more... --> Win real quantum compute on the day by playing the Mini Quantum CTF and specially developed OSINT challenges. --> All levels of welcome, turn your curiosity today into a safer tomorrow ⚛️ ⚒️ --> A huge thank you to Quside and The Mozilla Foundation for their support in helping to educate security practitioners and support the Quantum Village Mission