Every jargon term explained in a daily briefing ends up here, in plain language.
- AI gateway
- A middle layer companies place between their internal systems and an AI model, controlling what prompts and data flow to which model so the company isn't locked into one AI provider.
- AI agents
- AI software that goes beyond answering questions to independently carry out multi-step real-world tasks, like processing an email or approving a refund.
- CVE
- A unique ID number assigned to a specific software security flaw, letting security teams worldwide refer to and track the same vulnerability by a common name.
- MCP
- MCP (Model Context Protocol) is a common standard that lets AI agents safely access external tools and data.
- guardrails
- Safety rules built into an AI agent in advance so it can't take actions outside its approved scope.
- reinforcement learning
- A training method that rewards an AI for taking actions that get it closer to a goal. It's good at driving task performance, but it can also push the AI to prioritize completing the task over following safety rules.
- Digital Markets Act (DMA)
- An EU law designed to stop giant platforms like Google and Apple from abusing their market power. Violations can trigger fines of up to 10% of a company's global revenue.
- deepfake
- A fake image or video created by AI to realistically swap or synthesize a person's face or body.
- model weights
- The numerical values an AI model learns during training — they are, in effect, the model's actual knowledge and capability. Losing them means retraining from scratch with the same data and compute.
- model router
- A system that automatically picks a bigger, pricier model or a smaller, cheaper one depending on how hard a given prompt is — used to cut costs while keeping the performance a task needs.
- benchmark
- A standardized set of test tasks used to compare different AI models on the same yardstick — separate benchmarks exist for coding, reasoning, computer-use, and other skill areas.
- sandbox
- An isolated test environment set up so an AI or program can't affect real computers or networks outside it.
- spear phishing
- A targeted scam email attack aimed at a specific person or organization, as opposed to mass phishing sent to random recipients.
- safety classifier
- An automated screening system that flags risky or potentially abusive requests to an AI model, routing them to a more restricted model or blocking the response when needed.
- agentic AI
- AI that plans and carries out multiple steps on its own, without needing a human to direct every action.
- orchestrator
- A system that sits above multiple AI models and routes each task to whichever one fits best, rather than relying on a single model for everything.
- off-target effect
- An unwanted side effect in which a gene-editing tool cuts or alters the wrong spot in the DNA instead of its intended target.
- open-weight model
- An AI model whose trained parameters (weights) are published for anyone to download and run on their own servers. It differs from fully open-source software, which also releases the code, but it's often loosely called an 'open-source model' because it lets organizations run it independently.
- full-duplex voice
- A voice-conversation mode where both sides can listen and speak at the same time, like a phone call — instead of the stiffer take-turns pattern most voice assistants use.
- remote code execution
- A serious security flaw that lets an attacker run arbitrary commands on someone else's computer over the internet, without physical access.
- free cash flow
- The cash a company has left after paying for its investments and equipment, on top of its operating income. Negative means it's spending more than it's bringing in.
- capital expenditure
- Money spent on long-term equipment and assets, like data centers and servers — the kind of spending AI companies are pouring into infrastructure.
- autonomous AI agent
- An AI system that can set its own intermediate goals and carry out a chain of multi-step actions — searching, running code, manipulating files — without a human directing every step.
- alignment
- The area of AI research focused on making a model actually follow the goals and rules its developers and users intend, aiming to reduce the risk that a more capable model behaves in unintended ways.
- zero-day vulnerability
- A software flaw that its own maker doesn't yet know about, meaning it can be exploited by attackers before any fix exists.
- CRISPR gene editing
- A technology for precisely cutting or altering specific spots in DNA to edit genes, using a protein such as Cas9 as the 'scissors.'
- telemetry
- The automatically generated logs and monitoring data a system produces while running. When the volume gets enormous, spotting one anomalous signal inside it gets harder, not easier.
- token
- The small chunks of text (roughly word-fragments) that AI models process text in — usage fees for AI services are typically billed per token processed.
- frontier model
- Industry shorthand for the most capable AI models available at a given moment — the current state of the art, as opposed to smaller, cheaper models built for specific tasks.