Open Questions · 15 November 2026

Open Questions for 15 November 2026

After Duffy blamed a botched wire replacement for the FAA ground stop, will investigators confirm that account or find a software fault?

6 questions · 4 continued from earlier · about 12 minute read

The day's biggest stories, the question each one leaves open, and the desk's odds on each answer. The percentages are a judgment about the evidence so far, written by an AI model. They are not a measurement, and they are not advice.

First asked 14 November 2026

What triggered Friday's nationwide FAA ground stop: a botched line upgrade, a bad software update, failing old equipment, or an attack?

The FAA halted all U.S. flights for more than four hours on Friday, and the Trump administration now blames work to replace old copper communication lines.

The desk's odds on each answer
AnswerOdds14 November
Official findings confirm that contractor work replacing copper lines with fiber set off the failureThe Transportation Secretary, the FAA head and several outlets now give this account, and the contractor confirms the conversion work was under way. It is still a first-days explanation.60%new
Findings name a software or configuration change as the main triggerA reported FBI look at a software update keeps this alive. It rests on one outlet citing unnamed reporting, and officials have not said it.15%38% (−23)
Something else, or a mix that findings do not pin on one triggerThe router failure, the line work and old equipment may be reported as a chain with shared blame.14%14%
Failure or overload of aged communications equipment, with no outside triggerOfficials first stressed old equipment, but Saturday's account adds a specific trigger: the upgrade work.8%42% (−34)
Cyberattack or deliberate sabotageOfficials have ruled it out and no row reports evidence for it.3%6% (−3)

What moved the odds

Yesterday the desk split between old hardware and a software change. Duffy's Saturday account of a copper-to-fiber replacement that knocked out a router, and L3Harris's confirmation that the conversion is under way, moved most of the weight to a new upgrade-work answer.

What is known

What else has been going on

How questions like this usually turn out

Background knowledge: large outages in aviation and telecom systems usually trace to maintenance or a change made to a working system, not to an attack. First official explanations tend to hold in outline, while later reviews sharpen who or what made the error.

A caution on the coverage

The administration has reason to prefer an account that points to old equipment and a contractor, since it is asking for $50 billion to replace the system. L3Harris has reason to stress shared work with the FAA and telecom partners. The FBI software-update angle is reported second-hand and no official has confirmed it.

What would settle it

Expect to know: Two to four weeks, when written findings appear

16 sources

Will Microsoft's full review confirm that a network upgrade error caused its global outage, or point to something else?

Microsoft said Saturday it had restored Outlook, Teams, Copilot and other Microsoft 365 services after a global outage that disrupted businesses, schools and governments.

The desk's odds on each answer
AnswerOdds
Microsoft's detailed review confirms a faulty network capacity upgrade as the root causeThis is the company's stated cause and it fits how such outages usually happen. The rows carry no competing account.72%
The review names a different or additional main cause, such as a software defect or capacity shortfallMicrosoft was still hunting for the cause during the outage, so the first summary may be incomplete.15%
No detailed public root-cause report appears by mid-DecemberThe rows do not show Microsoft promising a date for a fuller account.10%
Microsoft or authorities later attribute the outage to an attack or intrusionMicrosoft has ruled this out and no row suggests otherwise.3%

What is known

What else has been going on

How questions like this usually turn out

Background knowledge: big cloud outages most often trace to a network or configuration change, and providers usually publish a review within weeks. The first stated cause tends to hold in outline, with later detail on why safeguards failed.

A caution on the coverage

Microsoft is the only source for the cause so far. It has reason to prefer an account of a routine upgrade error over one that suggests a security lapse or a deeper design weakness. Accounts of how long the outage lasted do not agree.

What would settle it

Expect to know: Within about a month

9 sources

First asked 14 November 2026

Will the U.S. carry out new military strikes on Iran in the next month?

Iran's Revolutionary Guard said tankers hit mines in the Strait of Hormuz and that it shot down a U.S. drone, while Saudi Arabia and Bahrain reported intercepting missiles and drones.

The desk's odds on each answer
AnswerOdds14 November
No new U.S. strikes on Iranian territory or forces by mid-DecemberNo row reports a U.S. response or a stated U.S. plan to strike. This is still the usual outcome in any given month.50%55% (−5)
Limited U.S. retaliatory strikes on IranThe mine reports and the claimed drone shootdown are the kind of incidents that have drawn narrow retaliation before. Both rest on Iranian statements.25%22% (+3)
U.S. strikes hit the Houthis but not IranHouthi missiles aimed at Riyadh give a nearer target if Washington chooses to act short of striking Iran.16%15% (+1)
A large-scale U.S. strike campaign resumesNetanyahu's language is strong, but it is Israel's position, and the rows show no U.S. move toward a wide campaign.9%8% (+1)

What moved the odds

The tanker mine reports, the claimed shootdown of a U.S. drone, the Bahrain intercept and Netanyahu's statement add pressure, so the no-strikes answer drops five points. The key claims come from Iranian statements and are not independently confirmed.

What is known

What else has been going on

How questions like this usually turn out

Background knowledge: after a major exchange of strikes, most single months pass without new direct attacks on a state's territory, even with harassment at sea. Mine incidents and drone shootdowns have at times drawn limited retaliation.

A caution on the coverage

Almost every account of the tanker fires and the drone comes from Iran's own military, and Reuters could not verify the first report. Iran has reason to show control of the strait, and Israel's government has reason to stress the threat. The rows do not say who laid the mines or how Washington reads the incidents.

What would settle it

Expect to know: By mid-December

12 sources

First asked 14 November 2026

Will Ukraine's government publicly accept withdrawing from any part of Donbas as part of a peace deal in the next month?

President Volodymyr Zelensky said U.S. negotiators floated the idea of Ukraine leaving Donbas in exchange for NATO membership, after talks with Russia in Abu Dhabi.

The desk's odds on each answer
AnswerOdds14 November
No acceptance; talks continue with Ukraine's position unchangedZelensky described the idea as a U.S. suggestion and did not endorse it. Nothing in the rows shows Kyiv moving.78%80% (−2)
Talks break down or are suspendedReported Russian demands about NATO and European objections widen the gap between the sides.9%7% (+2)
Ukraine signals conditional acceptance of a Donbas withdrawal tied to NATONATO membership is a long-standing Ukrainian goal and Zelensky says it is back on the agenda. That is not acceptance of withdrawal.8%7% (+1)
A referendum on a deal is formally announcedNo row reports any step toward a vote.5%6% (−1)

What moved the odds

Little moved. Zelensky's remark that NATO is back on the agenda shows the subject is live but not that withdrawal is accepted.

What is known

What else has been going on

How questions like this usually turn out

Background knowledge: governments at war rarely accept giving up territory they still hold within weeks of an idea being floated. Such proposals usually lead to more rounds of talks or to a stall.

A caution on the coverage

Zelensky is the only named source for what the U.S. proposed, and he has reason to make it public before the Shenzhen meeting. The Russian demands come from newspaper reports, one of which Moscow denies. Washington's own description of its position is missing from the rows.

What would settle it

Expect to know: By mid-December

15 sources

Will most states take the six-month grace period before enforcing Medicaid work requirements?

The Trump administration told states they may wait until July 2027 before removing people from Medicaid for failing to meet new work requirements.

The desk's odds on each answer
AnswerOdds
A majority of states say they will use the grace period by mid-DecemberThe delay costs a state little and lowers the risk of wrongly cutting people off. No row yet reports a state's decision.38%
Some states say they will use it, but fewer than halfSome states may prefer to enforce on the original schedule, and others may not decide within a month.34%
No reliable state-by-state count is published by mid-DecemberStates often take weeks to announce such choices, and the rows show no deadline for opting in.20%
The offer is changed or withdrawn, or Congress or a court alters the timelineThe policy is new and contested, so a further change is possible, though no row points to one.8%

What is known

What else has been going on

How questions like this usually turn out

Background knowledge: when Washington gives states an optional delay on a contested benefit rule, uptake tends to split along party lines, and states with unready computer systems tend to take the extra time. Many states decide slowly.

A caution on the coverage

The administration presents the delay as time for states to fix their systems, while critical coverage presents it as a pause on coverage losses. Because the grace period is optional, the national headline says little about what happens in any one state. The rows do not report any state's decision.

What would settle it

Expect to know: Mid-December for a first count

7 sources

First asked 14 November 2026

What will come out of the planned Trump-Xi-Putin meeting in Shenzhen?

The White House and the Kremlin say the U.S., Chinese and Russian leaders will hold their first three-way meeting in Shenzhen on Thursday.

The desk's odds on each answer
AnswerOdds14 November
Meeting held, with no joint statement or agreementThree governments with different aims must agree on wording within days. The rows show no draft text.45%44% (+1)
Meeting held, with a joint statement or announced start of arms-control talksAll three have reason to show a result from a first meeting, and a short statement is the easiest one to reach.34%32% (+2)
Meeting postponed, cancelled or not held with all three leaders in personBoth Washington and Moscow have confirmed the date, which lowers this risk as the day nears.9%12% (−3)
Something elseA separate two-way deal or a Ukraine announcement could become the main outcome.9%8% (+1)
A signed nuclear arms agreementNothing in the rows suggests a finished text is ready to sign.3%4% (−1)

What moved the odds

Little moved. Today's rows add a North Korean envoy's trip to Moscow and the first U.S.-China trade council meeting, neither of which bears directly on a three-way text.

What is known

What else has been going on

How questions like this usually turn out

Background knowledge: first meetings between rival leaders more often produce photographs and separate readouts than a joint text. Signed agreements usually need months of preparation that would be visible beforehand.

A caution on the coverage

Each government has reason to call the meeting historic whatever it yields. What the leaders will discuss comes from spokespeople, and the reported Russian demands rest on newspaper accounts. Separate readouts may describe the same meeting differently.

What would settle it

Expect to know: November 19 or the days after

12 sources

Still open from earlier editions

Not asked again today and not yet settled. The odds are the last ones given.

How this is made. Chosen and written by a model from the last 24 hours of the Corvus corpus, with up to 30 days of earlier reporting on the same subjects as background. The odds are the model's judgment, not a measurement. Links, outlets and earlier odds are joined from stored data, never written by the model. Written by claude-fable-5-2; no person edits it before publication.

Each statement names the outlets it rests on and links to their reporting. Open the source before relying on a detail. Earlier odds stay on the page for their date, so the record of what the desk said is not rewritten.