Пропустить навигацию
Without proper governance, an AI agent might autonomously access sensitive data, expose personal…
0:00 / 2:01

Without proper governance, an AI agent might autonomously access sensitive data, expose personal…

Andrew Ng

@AndrewYNg1,9 млн подписчиков

Открыть в X
951 отметка «Нравится»СкачатьОригинальный пост

78 560 просмотров

Without proper governance, an AI agent might autonomously access sensitive data, expose personal information, or modify sensitive records. In our new short course: “Governing AI Agents,” created with Databricks and taught by Amber Roberts, you’ll design AI agents that handle data safely, securely, and transparently across their entire lifecycle. You’ll learn to integrate governance into your agent’s workflow by controlling data access, ensuring privacy protection and implementing observability. Skills you'll gain: - Understand the four pillars of agent governance: Lifecycle management, risk management, security, and observability - Define appropriate data permissions for your agent - Create views or SQL queries that return only the data your agent should access - Anonymize and mask sensitive data like social security numbers and employee IDs - Log, evaluate, version, and deploy your agents on Databricks If you’re building or deploying AI agents, learning how to govern them is key to keeping systems safe and production-ready. Sign up here:

32 комментария

Комментарии из оригинального поста в X (Twitter)

  • Фото профиля Youssef El Manssouri

    @yoemsri

    @databricks No human employee gets root access on day one. Why would an autonomous system? Start with views Start with masks. Start with logs. Build trust through constraints, not hope.

    5 отметок «Нравится»

  • Фото профиля Neo

    @neo3141111

    @databricks Dear Sir, you are a legend 🙏 At the cusp of every era, you take the time to be the best teacher, a whole generation can look up to. A student of your ML lectures, and a current student of your Deep Learning lectures. ❤️ 🫡

    1 отметка «Нравится»

  • Фото профиля Sᴀɢᴀʀ kharal

    @skharal4_s

    @databricks Everyone's racing to ship AI agents. Almost nobody's talking about governance until something breaks in production. This is the unsexy work that prevents the viral incident.

    1 отметка «Нравится»

  • Фото профиля AIxBlock

    @AIxBlock

    @databricks Solid move. Agent governance isn’t a side topic, it’s what separates real AI systems from risky prototypes.

    1 отметка «Нравится»

  • Фото профиля Alpaca Network

    @AlpacaNetworkAI

    @databricks Essential work. Governance and transparency are what make AI sustainable long term. Open models and onchain infrastructure take that even further by making every action traceable and verifiable.

    1 отметка «Нравится»

  • Фото профиля Ronnel

    @ronney257

    @databricks Hey Mr. @AndrewYNg , can you please give me free access to your courses on Coursera, they are literally to expensive for me to afford, I'm from Nigeria and the price of the courses are way beyond my means of livelihood. I'm be eternally grateful

    1 отметка «Нравится»

  • Фото профиля Rohit Kumar Tiwari

    @_rohit_tiwari_

    @databricks This is such a relevant course. These skills are not optional but must.

    1 отметка «Нравится»

  • Фото профиля Michiel V

    @michielmv

    @databricks the governance gap is real - most people running agents locally are one bad skill install away from a problem. the tooling for this is still pretty immature

  • Фото профиля UncleChris1997

    @chui_chris69072

    @databricks 😎😎😎

  • Фото профиля Jimmy Louis

    @louis19_jimmy

    @databricks real build hours, no ghosting on data

  • Фото профиля AlphaHunterVN | ETHGas ⛽

    @VnAlpha78815

    @databricks The tension between true agent autonomy and necessary governance constraints is fascinating. What foundational guardrails are proving most effective?

  • Фото профиля Aiko Lang | AI-Powered Head of BD QStarLabs

    @aiko_qstarlabs

    @databricks AI governance isn't just a technical challenge - it's a human imperative. @AndrewYNg and @databricks show how we transform potential risks into responsible innovation. #AIEthics #ResponsibleTech

  • Фото профиля Anthara Fairooz

    @anthara_ai

    @databricks Great course on governing AI agents! Essential for data safety and transparency.

  • Фото профиля Michiel V

    @michielmv

    @databricks the lifecycle management piece is what most openclaw setups skip entirely. everyone focuses on building the agent, almost nobody thinks about governing it. signed up - this gap is real for anyone running persistent agents in production

  • Фото профиля THE AI REGULATOR

    @EUAIACTGUY

    @databricks 100%. Operationally: (1) tool scopes per role (least privilege), (2) approval gate for risky actions, (3) immutable audit receipts you can replay during incident response.

  • Фото профиля THE AI REGULATOR

    @EUAIACTGUY

    One governance pattern that prevents “agent went rogue” incidents: treat every tool like a production API. - explicit scopes per role (read vs write) - approval gate for high-risk actions (payments, deletions, sending emails) - immutable audit receipts (who/what/why + inputs/outputs) so you can replay an incident - continuous evals/red-team prompts after every model/tool change Without receipts + scopes, you can’t do incident response—only storytelling.

  • Фото профиля Sherif Negm

    @_chicoo

    @databricks

  • Фото профиля AI PlanetX

    @AI_PlanetX

    @databricks Great initiative! Governing AI agents is crucial for data safety and transparency. Excited to learn more!

  • Фото профиля Tobe Duru

    @tobe_duru

    @databricks governance is the new competitive advantage

  • Фото профиля TheAIGuy

    @TheAIGuy4u

    @databricks Critical insight on AI governance. As AI agents become more autonomous, what frameworks are you implementing to ensure responsible deployment while maintaining innovation velocity? #ResponsibleAI