America’s Got Governance

  • Coral Gables’ Art Cinema Bid: When the Runner-Up Gets Another Shot

    A recommendation, then a different route

    Coral Gables’ procurement record recommends Hycon Investments for the Art Cinema expansion contract, a project valued at roughly $1.24 million. The record describes Hycon as the lowest responsive and responsible bidder. But the recommendation was not the last step in the public story. The Coral Gables Gazette reports that on October 27, commissioners voted 5–0 to pursue West Construction if it would match or beat Hycon’s bid. In procurement, apparently, opening the bids does not necessarily close the plot.

    The question is about the process

    The Gazette reports that Hycon sued on October 30, challenging the city’s decision to give West another price opportunity after bids were opened. That is Hycon’s claim, not a court finding that the city violated its rules. The central question is whether the published solicitation allowed the approach commissioners chose, and how the city documented its reasoning. The city record recommending Hycon matters, but it is not a complete account of the commission’s later discussion or decision-making.

    A staff recommendation and a commission’s subsequent negotiating approach are separate steps. Residents should be able to see how the solicitation terms, bid tabulation, recommendation and commission action fit together. That paper trail is especially important when a bidder says the rules changed after bids were opened. Whether Hycon ultimately prevails is a separate question from whether the city has clearly explained its process.

    Keep the other facts in their own lanes

    There is more context, and it deserves careful handling. In follow-up reporting, the Gazette said West Construction president Matthew West was convicted in a federal fraud-related case. That reported conviction concerns the company’s president; it is not itself a finding about West Construction’s bid for Coral Gables or the city’s contract decision.

    The Gazette also reported campaign contributions to two commissioners who voted for negotiations. That is relevant disclosure information, but the reporting does not establish that the contributions caused or influenced the vote. Putting a conviction, contributions and a disputed procurement into one corruption-shaped ball would be easy—and would claim more than these facts show.

    What residents should be able to check

    The useful accountability test is more ordinary than a dramatic headline: can residents trace the city’s stated rules through the bids, staff recommendation and commission action? The lawsuit’s allegations remain unresolved, and the contribution reporting does not answer that process question. The city’s explanation and documentation do. Public procurement turns public rules into decisions about public money; when a bidder challenges how those rules were applied, the path from solicitation to award should be clear enough to follow without a decoder ring—or a complimentary fog machine.

    Key sources

  • |

    NHTSA Opens a Ford Explorer Airbag Probe. The File Is Not a Verdict.

    The investigation is open; the verdict is not. On November 6, 2026, the National Highway Traffic Safety Administration opened a preliminary evaluation concerning certain 2021–2026 Ford Explorers. The alleged issue is burn injuries following deployment of a seat-mounted side airbag. That is the concern described in NHTSA’s opening document—not a final finding that the vehicles have a defect. The filing cabinet has cleared its throat, but it has not issued a ruling.

    What the document counts

    NHTSA estimates that 1,236,155 vehicles are within the population under review. Its evaluation resume summarizes eight Vehicle Owner Questionnaires and four Early Warning Reporting death-and-injury reports. Separately, the document’s incident table lists 11 incidents, including 10 injury incidents. Those are different counts in the agency’s record, not figures to add together. They describe information NHTSA is reviewing; they are not a count of independently confirmed injuries, nor proof that every vehicle in the estimated population has the same problem. An estimate of scope is not a roll call of defective cars.

    What “preliminary” means

    The evaluation is meant to examine the issue’s scope, frequency, and potential safety consequences. NHTSA is assessing the reports and the information available to it; the opening document does not say that the agency has reached a conclusion about whether a safety defect exists. Nor does it announce a recall. Regulatory paperwork often arrives in stages: first the question, then the work of checking it, and only later—if the evidence supports it—a conclusion. The distinction is less dramatic than a recall headline, and more useful.

    Reports are leads, not verdicts

    Reuters and Anadolu also reported on the probe, describing it as preliminary and centering the reported burn-injury concern. The key word is “reported.” NHTSA’s incident table records information being evaluated; it does not establish that every case happened under identical circumstances or that the airbag caused each reported injury. The agency’s document is the source for the population estimate and incident tally, so those figures should travel with their attribution rather than wander around as settled fact.

    The practical meaning

    For drivers and families, the news is that federal safety scrutiny has begun—not that NHTSA has declared every Explorer in the population unsafe, or that a recall is underway. The estimate identifies the vehicles within the review’s scope; it does not certify them as affected. The evaluation is a step in examining a specific concern, a defined population, and reports the agency has received. It leaves open what those records will ultimately show. This is oversight before the paperwork reaches its final chapter: an agency asking what the reports add up to. Exhibit A has a pulse. The case is still open.

  • |

    The Pentagon Still Can’t Account for F-35 Parts Worth Millions

    The F-35 parts system is an inventory valued at approximately $31 billion, spread across more than 700 government- and contractor-managed facilities. It contains about 26 million parts. That is a lot of equipment, a lot of buildings, and apparently not enough reliable paperwork to tell the public what happened to previously reported losses valued at more than $85 million. The Pentagon can build a fighter jet that costs serious money; keeping a dependable inventory appears to be the part that keeps taxiing in circles.

    A very large inventory, a very basic problem

    In a report released November 9, the Government Accountability Office found that the Defense Department does not maintain complete, reliable records and physical inventories for F-35 parts. This is not a dispute about whether a spreadsheet looks tidy. If officials cannot trust the records or verify what is physically on hand, they cannot reliably oversee property spread across hundreds of facilities. The systems meant to track the parts are supposed to answer the boring questions—what is there, where is it, and what happened to it. Those are also questions taxpayers have a right to expect someone can answer.

    The most eye-catching figure needs careful handling. GAO reported losses valued at more than $85 million in a 2023 review. As of March 2026, the Joint Program Office could not determine the status of those earlier reported losses. That does not establish that the parts were stolen or that the full amount is currently missing. It establishes that officials could not determine what became of the reported losses. That is already a serious oversight failure without upgrading uncertainty into a crime story.

    The accounting fog has a second layer

    GAO also found that the program office could not determine whether a prime contractor’s approximately $28 million in submitted losses included any of the earlier losses. Read that twice before anyone starts adding figures in a cable-news graphics department: GAO did not establish that the $28 million was separate from the earlier amount, or that it entirely overlapped. The relationship could not be determined. When the official record cannot tell whether two piles of reported losses are connected, the public is left with numbers and no dependable map.

    GAO recommended stronger procedures for monitoring losses and improving property accountability. The Defense Department concurred. That is a useful first step, not proof that the problem has been fixed. A recommendation is paperwork with teeth only if the agency follows through and can show that its records and inventories now work.

    The takeaway is not that every unaccounted-for part was stolen. It is that a vast, expensive public program cannot provide a clear account of previously reported losses or reliably reconcile later submissions. Oversight depends on knowing what the government owns and what happened when something goes missing from the books. If the answer is still “we can’t determine that,” the public is paying for the aircraft and the fog machine.

    Sources

  • |

    Anthropic’s IPO Paperwork Puts Its AI-Governance Promises in the Public Filing Cabinet

    AI-governance promises have entered the investor paperwork stage. Anthropic’s amended S-1, filed November 10, describes how the company says it is supposed to balance its public-benefit purpose with commercial and competitive pressures. That is useful to have on the record. It is not the same as proving the machinery works when the pressure arrives wearing a revenue forecast.

    The structure Anthropic describes

    In the filing, Anthropic identifies itself as a Delaware public benefit corporation. It states its purpose as the responsible development and maintenance of advanced AI for humanity’s long-term benefit. The S-1 also describes the Long-Term Benefit Trust, which generally has authority to elect and remove a majority of the board. That is a governance design on paper: the Trust has a formal role in who oversees the company, rather than serving only as a slogan beside the company logo.

    But this is Anthropic describing its own structure and aims. A filing can tell investors what the company says its guardrails are; it cannot, by itself, demonstrate how those guardrails behave under stress. The existence of board authority identifies a formal source of power, but does not settle how that authority will be exercised in every difficult case. The distinction matters because decisions about AI safety can affect products and the people who use them, while the company also faces commercial objectives and competition.

    The tension is in the paperwork too

    Anthropic’s S-1 acknowledges that safety decisions can involve difficult judgments among safety considerations, commercial objectives, and competitive pressure. That admission makes the governance question more concrete. When those priorities pull in different directions, who has the authority—and the practical ability—to hold the line? The filing describes the Trust’s general board role, but it does not answer every future test in advance. No governance diagram can pre-file the awkward meeting where the incentives disagree.

    The proposed offering gives the filing a wider audience. The Associated Press reported on the proposed offering, but it remains proposed; the report is not evidence that an IPO or any particular outcome is settled. For prospective investors, the S-1 is part of the information they can assess. For users, the relevance is more basic: a company’s choices about safety and commercial priorities are not just internal housekeeping when its AI products are used in people’s work and daily lives.

    So the takeaway is neither “the Trust will save us” nor “the structure is meaningless.” Anthropic has disclosed a public-benefit purpose, described a Trust that generally has authority to elect and remove a board majority, and acknowledged that safety can involve tradeoffs with commercial and competitive goals. Those are concrete claims worth examining, not proof of successful accountability. The harder test is what happens when honoring the stated purpose has consequences. The filing puts the rules in the cabinet; accountability is what happens when someone has to use them.

  • |

    The Government Can Buy Data That Would Make a Warrant Fight Awkward

    Buying data is not opening your phone

    Government surveillance stories can arrive wearing a trench coat labeled “they can see everything.” The real policy here is narrower—and still worth scrutinizing. The Intelligence Community’s 2024 framework covers certain commercially available information: data sold or licensed by companies. That is not the same as officials simply accessing anyone’s phone. It is also distinct from information obtained through compulsory legal process. Those are different ways of obtaining information, not interchangeable descriptions of one power.

    The policy has a defined audience

    The framework sets principles for Intelligence Community elements accessing, collecting, and processing commercially available information. It does not automatically govern every federal law-enforcement agency. That scope matters. A document concerning government data use is not a universal permission slip stamped by the entire federal bureaucracy. The corkboard can connect dots; it cannot make separate agencies the same agency.

    Nor does the framework, by itself, settle whether a particular use of data is lawful under the Constitution or when a warrant may be required. The useful distinction is between what this policy says it covers and the wider legal questions that can arise around government access to information.

    Sensitive data, stated safeguards

    ODNI’s 2026 report discusses sensitive categories of commercially available information, reported uses, and safeguards the Intelligence Community says apply. The point is not that every product or purchase is identical; it is that “commercial” does not mean “harmless.” Data sold through business channels can still raise serious privacy questions when it is sensitive and used by government.

    The report offers a public account of the Intelligence Community’s stated approach. It does not establish that every possible purchase is appropriate, nor does describing safeguards answer every question about how data is selected, used, or overseen. A policy framework can make rules more visible without making the underlying data less sensitive.

    The transparency gap is the other story

    In a November 5 article, Lawfare argues that federal law-enforcement agencies have not provided comparable public frameworks or reporting on data buying. That is Lawfare’s assessment; it is not a finding that every agency operates without rules. But it raises a practical oversight question: if the public cannot readily see what data is purchased, how sensitive it is, and what limits apply, it is harder to assess whether the safeguards are meaningful.

    Follow the thread, check the knot

    So the 2024 framework does not say officials can casually browse everybody’s phone. And the existence of a commercial data market does not make government use automatically benign. The important questions are more specific: what information can agencies buy or license, how is it used, what safeguards apply, and what can the public inspect? ODNI’s documents provide a window into the Intelligence Community’s stated approach. Lawfare says the public view is much less clear for federal law enforcement. Ordinary people deserve more than a rumor with a ring light on one side and an unmarked policy binder on the other.

  • |

    The DHS Contracting Questions Following Corey Lewandowski to DOJ

    Sen. Maggie Hassan renewed her request for Department of Homeland Security records about Corey Lewandowski on November 10, putting contracting oversight back in focus. Her announcement follows reporting that federal prosecutors are investigating allegations concerning Lewandowski’s time at DHS. That is a serious public question, not a verdict. In Washington, the paperwork can be the most boring character in the story—until it helps show who had authority, what rules applied, and whether oversight went beyond a box checked in triplicate.

    What has been reported

    ABC News and Reuters reported on October 22 that the Justice Department was investigating allegations related to Lewandowski’s DHS role and contracting. ABC reported that the inquiry followed a referral from the department’s inspector general. That sequence is worth keeping precise: the reporting describes an investigation and a referral, not charges, an indictment, or an official finding that wrongdoing occurred. The distinction is not a technicality. It is the line between scrutiny and a conclusion supported by evidence.

    ABC also reported that Lewandowski’s lawyer denied wrongdoing and said the Justice Department had not contacted him. That statement matters too. A reported investigation does not publicly establish what happened, and a denial does not settle the questions either. The central details remain unresolved in the reporting cited here. The money-trail correspondent would love a neat receipt; for now, the public has reporting about an inquiry and a denial, not a completed account of the facts.

    Why Hassan wants the records

    Hassan’s November 9 letter, followed by her November 10 announcement, renews a request for records concerning Lewandowski and his DHS role. The letter documents what a senator is seeking; it is not independent evidence that the allegations are true. The request may help clarify what information is available about contracting oversight, but its contents have not been established by the request itself. The announcement also does not mean DHS has turned over the material or that the public now has the answers.

    A request and a response are two separate pieces of paperwork, and neither should be mistaken for the final accounting. The useful question is whether records and any eventual official findings clarify who was responsible for oversight and what the process showed. Taxpayers have a stake whenever public contracting is under scrutiny: public agencies spend public money, and the public deserves enough transparency to understand how decisions were reviewed.

    Until evidence fills in the gaps, keep the verbs honest: Hassan requested; news outlets reported; prosecutors are investigating. Scrutiny is not proof, and a demand for documents is not the documents themselves. The conclusion still has to earn its signature.

  • |

    The Interior Secretary’s AI Investments Meet Public Land

    When the form is broad and the policy is public

    A disclosure may satisfy filing rules and still leave the public asking what, exactly, is in the room. Interior Secretary Doug Burgum’s reported financial interests overlap with his role in advancing data-center development on federal lands. That overlap does not establish improper influence. It does make the paperwork worth reading slowly, preferably with a lamp and the uneasy feeling that a filing cabinet has just cleared its throat.

    What the estimate does—and does not—say

    ProPublica reports that Burgum’s 2026 financial disclosure lists interests in Arthur Ventures, the investment firm he co-founded. ProPublica estimates that $18.1 million to $64.4 million of those investments are in AI companies. That estimate is based on Save Our Parks research, which examined company marketing materials; it is not a precise amount stated on the disclosure form. The form gives broad value ranges. Those are meaningfully different things: a disclosed range is not an itemized account, and an outside estimate is not independently verified accounting.

    The distinction matters because a large-looking estimate can travel farther than the qualifications attached to it. The available record, as described in the reporting, does not turn that range into an exact inventory of Burgum’s AI holdings. The point is not to pretend the estimate says nothing; it is to keep clear what kind of number it is and where it came from.

    The public-interest overlap

    The policy question is concrete even if the financial picture is not exact. Burgum leads the Interior Department as it advances data-center development on federal lands, while the reporting identifies investment interests connected to AI companies. Interior says Burgum complied with ethics and disclosure requirements and that his decisions were not driven by personal investments. That response matters. So does the distinction between compliance and confidence: the department’s position does not, by itself, answer every public question about how decisions affecting public land are reviewed.

    A USGS report on data centers and Bureau of Land Management lands discusses considerations including energy, transmission and water. Those are practical land-management issues, not proof of a conflict. They are also reasons the public may reasonably want to understand what is proposed and how the agency weighs competing considerations. The review process should be legible in the records, not left to inference from a reassuring paragraph or a broad financial-disclosure category.

    The next document matters

    On November 6, Sen. Martin Heinrich sent BLM a letter seeking information about the agency’s authority, proposals, locations, acreage, and plans for environmental review or public comment. That is an oversight request, not a finding of wrongdoing—and it is not evidence that any proposal has become a completed project. It puts a practical accountability question on the record: what information will BLM make available about proposed data centers and the process for reviewing them?

    Public land belongs to the public in more than a ceremonial sense. People should be able to see what is proposed, what is being weighed, and how decisions will be made. The disclosure and Interior’s response are part of the record; so is the senator’s request for more detail. At this stage, the important question is what information the agency provides next.

    Key sources

  • |

    The Election-Fraud Group Chat Kept Posting as Results Emerged

    A rumor gets a head start

    As midterm results emerged, a claim that an FBI raid in Houston had uncovered ballots began circulating online. The Associated Press reported that the claim was fabricated. AP also cited University of Washington researchers who observed a burst of X posts about malfunctioning voting machines the following morning. Those posts showed claims moving around online; they did not establish that fraud had occurred. The distinction is less exciting than a corkboard covered in red string, but considerably more useful.

    Circulation is not a verdict

    AP characterized false accusations of election fraud as limited. That qualification matters. A loud cluster of posts can feel like a national emergency when it arrives in the same feed as breaking results, angry commentary, and a cousin typing “just asking questions” with both thumbs. But volume in a feed is not a measure of how many people believe a claim, and repetition does not turn a claim into evidence. The algorithm may wear a trench coat; it still cannot certify a rumor.

    NewsGuard’s 2026 midterms misinformation tracker offers another specific measure, with important boundaries. As of November 5, it recorded 55 false midterm claims and 611 confirmed appearances across platforms among the claims it tracked. NewsGuard said 64% of those appearances were on X. These are the tracker’s documented instances—not a count of every false claim online, every post, or every platform interaction. Think of it as a careful tally from one defined clipboard, not a census of the internet wearing a fake mustache.

    Labels can help, but absence proves nothing

    The Atlantic has reported on limits in X’s Community Notes system for polarizing election claims. Some claims may not receive a published note. That gap is not evidence that a claim is true, just as a note appearing is not a substitute for checking the underlying evidence. Nor did the absence of a note cause the Houston rumor or the machine posts. Notes are one tool in a messy information environment, not an all-seeing referee with a whistle and a perfect signal.

    For people trying to follow results, it helps to separate three things: what a post alleges, what reporting has verified, and how widely the claim appears to be circulating. Here, AP identified the Houston story as fabricated, described the machine-related posts as online activity, and said false fraud accusations were limited. NewsGuard’s numbers describe its own tracked claims and appearances. That is enough to see the pattern without inflating it into a sweeping fraud wave. Rumors can gather momentum because they offer instant certainty and a ready-made group chat. The evidence still has to do the slower work. Follow the thread, sure—but check the knot.

  • |

    Raytheon’s Radar Deal: A Nearly $3 Billion Ceiling and a Very Long Memory

    The Missile Defense Agency has extended Raytheon’s AN/TPY-2 radar operations and sustainment contract for two years, through October 31, 2028. The noncompetitive modification raises the contract ceiling by $394 million, from $2.605 billion to $2.999 billion. The amount obligated at award was $112,309,403. Those are three different numbers doing three different jobs, which is apparently why federal contracting keeps a calculator near the door.

    A ceiling is not a check

    According to the Pentagon’s October 21 contract announcement, whose terms The Daily Beast also summarized, the modification extends the radar-support arrangement. The nearly $3 billion figure is the maximum ceiling after the increase—not a report that the government just paid Raytheon that amount. Nor does the $394 million ceiling increase mean all of that money was newly obligated. The $112,309,403 is the amount identified as obligated at award. Taxpayers deserve the distinction because “contract value” can sound like money already out the door when it is not.

    The AN/TPY-2 is a missile-defense radar, and the contract covers its operations and sustainment. Continuity can matter for specialized defense work: an agency may have practical reasons to extend an existing arrangement rather than interrupt support. But “noncompetitive” still deserves a plain-language explanation from the agency. What alternatives were considered, and how did officials weigh continuity, cost, and contractor performance? Those are reasonable oversight questions—not an accusation that this extension was improper.

    Past conduct is context, not a verdict on this award

    Raytheon’s corporate history makes the oversight question harder to ignore. In 2024, the company reached a resolution with the Justice Department involving allegations of defective pricing on government contracts, foreign bribery, and export-control violations. DOJ announced that Raytheon agreed to pay more than $950 million in connection with the matters. The SEC-filed deferred prosecution agreement documents the company’s accepted responsibility under that agreement.

    That resolution concerns past conduct. It is not a new finding of wrongdoing about the AN/TPY-2 extension, and it does not establish that this contract action was tainted. Keeping those facts separate is not corporate courtesy; it is how accountability stays attached to evidence instead of turning into a rumor wearing a badge.

    Follow the invoice—and the oversight

    The public-interest question is whether the Missile Defense Agency can explain why a noncompetitive extension is the right way to keep this radar support running, and how it will monitor performance and spending under the modified contract. A large ceiling is not automatically a large payment, and a prior corporate resolution is not automatic proof of a new offense. But neither fact should end the conversation. Public money comes with a public duty to show what was obligated, what work is being delivered, and how the agency oversees the contract. The invoice should come with footnotes.

  • |

    DOJ’s Informant Rulebook: Recommendations Agreed to, Reforms Still Pending

    The federal bureaucracy has a phrase for the moment when everyone agrees the paperwork should change: “recommendation concurred with.” It is not the same phrase as “change completed,” though the two can look cozy in a status table. In an audit released October 28, 2026, the Justice Department’s Office of Inspector General reviewed confidential-informant policies at ATF, DEA, the FBI and the U.S. Marshals Service. The report’s central lesson is a small but sturdy one: agreement is a promise; evidence is the receipt.

    Policies reviewed, not every operation

    The OIG found the four components were in substantial compliance with applicable guidance. That is important context, not a footnote to be hurried past. The audit reviewed policy content and related controls; it did not establish that every field operation followed or violated policy, and it did not make a broad finding of operational misconduct. The document is careful about its boundaries. The filing cabinet cleared its throat, then stayed inside the lines.

    Within that generally compliant picture, the auditors identified areas needing attention: policies should account for emerging risks, internal controls were not consistent across components, and guidance needed to address electronic payments. The report also highlighted the monitoring of informants authorized to engage in otherwise illegal activity. Those are policy and oversight concerns—not a finding that agents or informants broke the rules. The distinction matters when the subject is confidential sources and the public record can easily grow more dramatic than the evidence.

    Concurrence is not closure

    The OIG made five recommendations, and DOJ and all four components concurred with them. The recommendations address the report’s identified policy and control gaps, including emerging risks, inconsistent controls, electronic payments and oversight of informants authorized to engage in otherwise illegal activity. The agencies’ agreement is not evidence that the reforms are already finished.

    That is where the administrative fog tends to gather. The OIG describes the report as resolved, but that status is not the same as closing the recommendations. The OIG says it will close them only after receiving evidence that the specified changes were made. “We agree” is a useful first entry in the ledger. It is not the final stamped page.

    For the public, this is the unglamorous part of accountability that actually counts. A policy can be sensible on paper, and an agency can accept an auditor’s advice, but oversight still needs proof of implementation. The audit does not claim the informant system broadly failed. It identifies places where rules and controls need strengthening, then leaves the recommendations open to the evidence-and-closure test. That distinction keeps a promise from being mistaken for a completed repair. In government, the document coughed. The next question is whether anyone brings the receipt.

End of content

End of content