Sep 1 – Oct 1, 2026 · 5 items
The AI week, for a policy adviser
A real edition, written for: Policy Adviser
The one thing
Dots can now run 24/7 on briefing updates—could this replace your current workflow?
Dots launch: always-on assistants for ongoing tasks
OpenAI launched Dots, always-on AI assistants that work without constant direction. They run on GPT-6 Astra with their own cloud computer. Dots connect to over 4,000 apps via plugins. They learn user preferences and work 24/7 on tasks. Users can inspect a dot’s work on its cloud computer. Dots can also use a user’s laptop with permission. They message progress or questions in ChatGPT, Slack, or Teams. Built-in safeguards include read-only tools for background work. Users control access, permissions, and approvals via Custom Rules. Specialist dots for organizations have their own identity and access. These are in focused enterprise pilots. Microsoft Agent 365 will integrate specialist dots. Dots roll out to Pro and Business Premium users in eligible markets.
Why it matters for you
You draft briefings that need constant updates. Dots now run 24/7 on tasks like this.
- Pro and Business Premium plans get first access
- They message progress or questions in ChatGPT
- Built-in safeguards include read-only tools for background work
Try this
Set up a Dot to monitor and summarise daily policy updates for your briefing
Paste this into your AI tool
Context: I am a Policy Adviser drafting briefings for senior officials in a Swiss government department. Task: Monitor and summarise daily policy updates relevant to our department’s current priorities. Constraints: - Only use public, official sources - Summarise in bullet points under 20 words each - Flag any urgent or high-impact items I'll paste: [today’s policy update sources] Output: A concise daily summary with clear action points for senior officials
Sources
Perplexity Automations for long-running agents
Perplexity Computer launched Automations for long-running agents. Agents remember prior work and pick up where they left off. They run on a schedule or when events occur in Slack, Gmail, Outlook, Linear, or GitHub. Users set triggers, define tasks, and specify outputs like Slack or Gmail. Automations can flag missed updates or deadlines. Users create them via the omnibar or the Automations interface. Admins can limit connector permissions. Automations do not use credits while waiting for triggers. Existing Scheduled Tasks can be migrated to Automations. Automations are available to all Computer users.
Why it matters for you
Your briefings often pull from long threads. Automations now remember prior work and trigger on events.
- Agents pick up where they left off after interruptions
- Triggers work from Slack, Gmail, Outlook, Linear, or GitHub
- No credit use while waiting for triggers
Try this
Ask your IT team whether Automations can replace your current briefing update workflow
Cisco Talos tracks AI-integrated malware
Cisco Talos built an open-source framework called CAIRN. It flags AI-integration traits in malware to classify and track samples. CAIRN already found a Windows hacking tool named CLOSEDQUORUM. CLOSEDQUORUM polls DeepSeek, Qwen, Mistral and Google Gemini for directives. The tool has no human input mechanism. It steals login credentials and cryptocurrency. Researchers linked it to cybercriminal forums from 2025. They could not confirm its real-world use. CAIRN also uncovered about 20 other AI-integrated malware examples. Researchers say AI is becoming operationalized for attackers.
Why it matters for you
Government systems are high-value targets. This framework flags AI traits in malware to classify threats.
- CAIRN found 20 AI-integrated malware families
- One tool polled DeepSeek, Qwen, Mistral and Google Gemini for directives
- Researchers say AI is becoming operationalized for attackers
Try this
Check with your cybersecurity team if CAIRN is on their radar
Sources
Skild AI: robots surpass human capability via self-play
Skild AI announced a breakthrough in physical AI using self-play. A strong base model like S1 can now learn complex tasks such as soccer by competing against itself in simulation. This approach allows robots to surpass human capability. The model started unable to walk but learned to dribble, shield, and tackle over time. After 140 years of simulated play, the policy transferred successfully to a humanoid robot. Soccer was chosen as a testbed for its physical and strategic demands. The method is general and will extend to other robotics tasks. Skild plans to scale this to larger teams and new domains. The work revives self-play as a path to physical AGI. The results build on Skild’s S1 robotics foundation model.
Why it matters for you
Policy on emerging tech needs concrete examples. This shows robots learning complex tasks beyond human skill.
- A strong base model learned soccer via 140 years of simulated play
- Policy transferred successfully to a humanoid robot
- Method is general and extends to other robotics tasks
Try this
Note this as a case study for your next briefing on AI and automation
Sources
NVIDIA launches Open Agent Safety Platform
NVIDIA launched Open Agent Safety Platform for full-stack agent governance. It combines OpenShell software and Sentry reference design. OpenShell sets secure runtime boundaries on NVIDIA Vera CPUs. It can extend to Arm and Intel platforms. Sentry runs on BlueField-4 DPUs to monitor agents in real time. Sentry quarantines agents breaking boundaries in milliseconds. Over 100 organizations are collaborating on the platform. Partners include Anthropic, Cisco, CrowdStrike, Dell, Figure, HPE, Hugging Face, JPMorgan Chase, Microsoft, Palantir, Red Hat, Salesforce, SAP, Scale AI, ServiceNow and SpaceXAI. The platform is available now via NVIDIA’s developer resources and GitHub. It supports the Open Secure AI Alliance mission.
Why it matters for you
Agent governance is now a hardware-level concern. This platform quarantines agents breaking boundaries in milliseconds.
- Sentry runs on BlueField-4 DPUs to monitor agents in real time
- Over 100 organizations are collaborating on the platform
- Partners include Anthropic, Microsoft, and Palantir
Try this
Ask your CTO if this changes your department’s agent deployment plans
Sources
Build this
Every week, one small thing to build with AI in something you actually care about. No work in it. Five minutes to set up, and worth keeping if it earns a second run.
A ranked list of three policy options, with the weakest one defended.
Five minutes to set up
Context: I am evaluating policy options for a Swiss government briefing. Task: Rank these three options from strongest to weakest and justify each rank in one sentence. I'll paste: [option A] I'll paste: [option B] I'll paste: [option C] Output: Ranked list with one-sentence justification for each
- Paste your three policy options into the prompt above
- Run it in ChatGPT and note the ranked order
- Now ask it to defend the option it ranked last in one paragraph
- Save the sequence as a reusable workflow for future briefings
Yours arrives Thursday.
This one was written for a policy adviser. Tell us what you do and the next one is written for you — same news, your job, once a week.