Tech news, 10 November 2026: Claude Fable 5.2 and Mythos 5.2, N-able takes N-central offline, Qubole's AWS access abused, Norway pauses data centre grid bookings and Meta's Muse Ember

1. Anthropic releases Claude Fable 5.2 and Mythos 5.2

Anthropic released Claude Fable 5.2, available to all users and on major cloud platforms, and Claude Mythos 5.2, restricted to vetted organisations, on 9 November¹. Anthropic says both produce output about 30% faster than their 5.1 predecessors, and Fable 5.2 is priced at $8 per million input tokens and $40 per million output tokens, down from $10 and $50¹².

CTO take: a price cut on a top model changes which tasks are worth routing to it. Re-run your evaluation sets on the new model, compare cost per completed task rather than per token, and only move traffic where results hold. If switching models needs code changes, fix that first; see model routing and evaluation sets that predict production.

2. N-able takes N-central offline after alleged zero-days are published

N-able took its hosted N-central remote monitoring and management environments offline in the Americas, Europe and Asia-Pacific on 9 November, after a researcher published proof-of-concept exploits for more than 30 alleged zero-day flaws; on-premises customers were told to take servers down until further notice³. N-able said it was developing a hotfix; details of affected versions had not been published at the time of reporting³⁴.

CTO take: remote management tools hold the keys to every machine they manage, which makes them some of the most valuable targets in any estate. If your managed service provider uses N-central, ask what they have done, how they will check for compromise and when they expect to restore service. More broadly, know which suppliers can run commands on your machines, and make sure you could cut that access quickly. See the MSP cyber resilience case study.

3. Attackers use Qubole's trusted access to reach customers' AWS accounts

Wiz reported that an attacker used credentials tied to Qubole's AWS account to assume integration roles that customers had created to trust Qubole, then collected secrets and downloaded data, between 14 and 26 October⁵. How the attacker obtained the credentials has not been disclosed, and the number of affected organisations is not known⁵.

CTO take: every cross-account role you grant a vendor is a door into your cloud that the vendor's security protects. List the roles that trust outside accounts, remove ones for services you no longer use, narrow what each can read, and add conditions such as expected source networks where the vendor supports them. Keep secrets out of places integration roles can read. See the machine identity case study.

4. Norway pauses new grid reservations for data centres

Norway's grid operator Statnett said on 9 November that it is temporarily halting new grid capacity reservations for data centres; projects with approved requirements can still connect and existing connections are unaffected, DCD reports⁶. The government is expected to introduce a licensing scheme that favours data centres serving critical needs⁶.

CTO take: Nordic countries have been a default answer for cheap, clean power for compute. That is no longer a given. If your plans assume new capacity in Norway, confirm your provider's projects are already approved, and keep options in other regions. See planning around the AI capacity crunch and the Europe page.

5. Meta launches Muse Ember, a small model for agent tasks

Meta released Muse Ember, its smallest and cheapest model, aimed at repetitive agent tasks such as tool calls and document parsing, priced at $0.10 per million input tokens and $0.50 per million output tokens, The Next Web reports⁷. Meta's comparisons with competing small models come from its own tests⁷.

CTO take: most agent workloads are many small, repetitive steps. Routing those to small models and saving large models for hard reasoning is where the biggest savings sit. Test small models on your own tasks before relying on vendor benchmarks.

What I'd do this week

• Re-run evaluations on new model versions before switching traffic.
• Ask your managed service provider whether they use N-central and what they have done.
• List AWS roles that trust outside accounts and remove or narrow them.
• Check where your future compute capacity is planned and whether it is approved.

I am based in India, the United States and the UAE, working globally. See where I work, or analyse your challenge and bring the result to a call.