EMET NEWS PRESS

Why does the public record not support a verified series?

The starting point is not a completed investigative finding, but a publicly circulating case list. It connects individuals who, at different points, had contact with US research, space systems, nuclear infrastructure, defence structures, materials science or UAP-related narratives.

The House Oversight Committee took up the issue in April 2026 and requested information from NASA, the FBI, the Department of Energy and the Department of Defense. The committee referred to unconfirmed public reporting. That is a political request for review, not proof of linkage.

CBS News reported that the FBI was examining possible connections between several deaths and missing-person cases involving people connected to government laboratories, nuclear technology or space technology. The category remains broad: scientists, former military personnel, administrative staff, technical employees and contractors are grouped together despite materially different functions, access levels and event circumstances.

The central error in many public accounts is the conflation of institutional proximity with operational relevance. Employment in a security-adjacent environment does not establish access to sensitive programmes. A technical research context does not establish covert access.

A case becomes analytically robust only when common programmes, timelines, perpetrator profiles, access patterns, contact traces, data exfiltration or foreign-state use can be identified. These indicators are not publicly present at this stage.

This limitation is central: not publicly established does not mean excluded. In a counterintelligence context, relevant findings may be classified, reserved for active investigations or operationally non-disclosable. OSINT cannot conclusively negate a covert situation. It can only assess whether open sources support a defensible pattern. That pattern is not currently demonstrable.

Which individuals are central to the case list?

Nuno Loureiro was Professor of Nuclear Science and Engineering and Physics at MIT and director of the Plasma Science and Fusion Center. His work covered plasma physics, magnetic reconnection, fusion research and astrophysical plasma processes. The profile was scientifically senior and technology-politically relevant.

The killing of Loureiro is a serious criminal case, but open-source reporting does not substantiate foreign-state access. The FBI assessed Claudio Manuel Neves Valente as a lone actor and found no terrorism or wider conspiracy nexus.

The known investigative picture points to personal grievance, not an access pattern against fusion research. Operational assessment: technology-politically relevant, but not a robust series indicator.

Carl Grillmair was an astrophysicist at Caltech/IPAC. His profile covered galactic structure, stellar streams, exoplanet research and NASA-adjacent observation programmes. Work linked to NEOWISE and NEO Surveyor touches planetary defence, but does not publicly establish classified weapons, propulsion or intelligence research.

His death is being prosecuted as a homicide. Freddy Snyder has been charged with murder. The known allegations include murder, carjacking and burglary.

The event profile points to local violent crime, not access against space programmes. Operational assessment: scientifically relevant, weak for the series hypothesis.

William Neil McCasland is the most security-relevant name on the list. He was a retired US Air Force major general, former commander of the Air Force Research Laboratory and connected to military research, space acquisition and Special Programs. AFRL, Wright-Patterson, Kirtland and Space Acquisition sit in the core area of advanced military technology.

McCasland disappeared from Albuquerque on 27 February 2026. Reports describe electronic devices, glasses and wearables left behind; hiking boots, wallet and a revolver were reportedly missing. His wife rejected UAP speculation and stated that he had been retired for more than a decade and no longer held current classified access.

McCasland remains operationally reviewable because of his former role. Context knowledge from earlier programmes may remain exploitable after retirement, especially regarding programme architecture, acquisition logic, technological development paths, institutional networks and Special Programs procedures. Retirement reduces the likelihood of current detailed knowledge; it does not automatically eliminate counterintelligence value.

For abduction, defection or foreign-state access, public indicators are absent: travel movement, contact trail, demand signal, claim of responsibility, technical compromise, adversary use or access signature. Operational assessment: highest security relevance, but no publicly established access finding.

Monica Jacinto Reza was a materials scientist and metallurgist connected to NASA JPL and rocket-engine materials. Public profiles identify aerospace, metallurgy, materials science and superalloys as her fields. Nickel-based superalloys for high-temperature applications are dual-use relevant because they involve engines, thermal protection, high-temperature components and space propulsion.

Reza disappeared on 22 June 2025 while hiking in the Angeles National Forest. The case remains open and is therefore security-reviewable. A disappearance in difficult terrain without a known contact, demand or foreign-linkage profile does not establish abduction.

Reza is relevant because of her technical profile, not UAP narratives. The defensible review basis is materials science, space propulsion and dual-use competence. Operational assessment: technically serious, but without a publicly established link between her work and disappearance.

Melissa Casias worked at Los Alamos National Laboratory. She disappeared in June 2025 and was found dead in Carson National Forest in late May 2026. Reports described her as an administrative assistant or laboratory employee. The cause of death had not been conclusively reported at the time; a handgun was found near the remains.

The security amplification comes from Los Alamos. LANL is a core institution of the US nuclear research complex. Affiliation with LANL does not establish high-level access or programme responsibility.

Casias is therefore a case of institutional proximity, not established technical target relevance. Operational assessment: reviewable death in the LANL environment, but no robust nuclear-scientist or access finding.

Anthony Chavez is listed as a Los Alamos case. Publicly available accounts describe him as a retired construction engineer or former construction foreman at Los Alamos National Laboratory. That indicates site and infrastructure exposure, but not a documented profile as a nuclear scientist or programme owner.

His disappearance in May 2025 remains unresolved. Age, retirement and the absence of public foreign-interference indicators leave the case open, but not automatically counterintelligence-relevant.

Infrastructure knowledge can matter. An access hypothesis would require indicators of contact, abduction, data leakage or foreign linkage. These indicators are not publicly available. Operational assessment: infrastructure-adjacent and security-relevant in a limited sense; only partly probative for the series hypothesis.

Steven Garcia disappeared in August 2025 in Albuquerque. Reports describe him as a government contractor or property custodian connected to the Kansas City National Security Campus or an NNSA-adjacent structure. The site is relevant to non-nuclear components of the US nuclear weapons complex.

The function property custodian does not establish scientific programme authority. It may imply access to buildings, material inventories or internal procedures; it does not automatically imply knowledge of strategic weapons designs or exploitable programme content.

Garcia is not publicly established as a nuclear scientist. The case remains reviewable because of site affiliation, NNSA proximity and missing-person status. Operational assessment: security-adjacent open missing-person case; no confirmed counterintelligence picture from open sources.

Jason Thomas worked at Novartis in chemical biology. He was reported missing in December 2025 and found dead in Massachusetts in March 2026. Reports point to personal strain; authorities did not assess foul play.

Thomas does not fit a nuclear or space-technology frame. His field was pharmaceutical-biotechnological, not publicly military-strategic. His inclusion demonstrates the expansion of the selection frame: “scientist” and “sensitive research” are treated as sufficient even when the technical domain does not align.

Operational assessment: analytically weak for the series hypothesis.

Michael David Hicks was a planetary scientist and JPL researcher focused on comets, asteroids and near-Earth objects. This touches planetary defence, early warning and the space situational picture.

His death in July 2023 predates the current public compression. There is no public evidence of foul play or security context. His inclusion rests on NASA/JPL proximity and NEO research, not on a known access finding.

Operational assessment: NASA-adjacent and technically relevant, but temporally distant and operationally weak.

Frank Maiwald was connected to NASA JPL and worked in space instrumentation and imaging spectroscopy. Sensors, spectroscopy and remote sensing can be dual-use adjacent.

Public obituaries report his death in July 2024; a confirmed cause of death was not widely reported. There is no robust public indication of foul play. Without an investigative finding or evidence of external action, the case is not operationally probative.

Operational assessment: technically relevant, but no access finding.

Amy Eskridge occupies a separate category. She is primarily linked in public narratives to UFO, anti-gravity and exotic-technology claims. Her profile is methodologically weaker because open sources show less robust institutional anchoring, no clearly documented access to classified programmes and no firm official placement inside a security-relevant case group.

Her function in the narrative emerges mainly through UAP, anti-gravity and “suppressed technology” references, not through a documented programme or access finding. Operational assessment: publicly resonant, but not robust for the series hypothesis.

What actually connects the cases?

The common denominator is limited.

Several individuals had proximity to US high-technology institutions: NASA JPL, LANL, MIT, Caltech, AFRL and NNSA-adjacent facilities. That proximity is real, but uneven.

Not all were scientists. Casias was an administrative assistant, Chavez a former construction foreman, Garcia a property custodian or contractor. Such roles may be security-adjacent, but they are not scientific key functions.

The cases do not sit in a tight timeline. They range from 2022 to 2026. A counterintelligence-relevant series would require closer temporal, methodological or target correlation.

The event types differ materially: homicides with suspects, open missing-person cases, deaths without known foul-play findings, personal-crisis contexts and unresolved cases. This heterogeneity weighs against a robust common-control assessment.

The base-rate argument explains why an apparent cluster does not automatically constitute an operational pattern: the United States employs thousands of people in nuclear, space, defence, laboratory and high-technology contexts. In that population, deaths and missing-person cases are statistically expected.

This argument is a methodological caution, not a counterproof. Without hard base rates, it cannot exclude a genuine early-stage pattern. It only blocks the premature inference from retrospective clustering to operational control.

Which cases remain security-relevant?

Four cases remain operationally reviewable.

McCasland because of former AFRL, Space Acquisition and Special Programs functions.

Reza because of materials science, superalloys and dual-use relevance for space and propulsion technology.

Garcia because of NNSA-adjacent infrastructure and possible site access.

Chavez because of LANL infrastructure exposure, with limited scientific value.

These cases justify security review. The public record does not currently support a robust series finding. Counterintelligence reviews whether a pattern exists. Journalism must not write as if that pattern were established.

Where is the analytical failure in many public accounts?

The “missing nuclear scientists” narrative creates three distortions.

First: professional profiles are upgraded. Administrative, technical or infrastructure roles become “nuclear scientists”.

Second: timelines are compressed. Cases across several years appear as a current situation.

Third: evidentiary burdens are shifted. Lack of explanation is treated as evidence of covert control. That is analytically invalid. Unknown is not proof.

 

What counterintelligence assessment is defensible?

Foreign-state collection against US science and high technology is structurally plausible. Russia, China, Iran and other actors conduct scientific, technological and procurement intelligence. Target fields include space systems, sensors, materials science, nuclear technology, fusion, semiconductors, AI and propulsion.

Structural plausibility does not replace case-specific evidence. A defensible operational hypothesis would require shared target programmes, access methods, digital or physical pre-operational activity, foreign linkage, contact chains, data leakage, travel movements, financial traces, perpetrator profiles, communication patterns or exploitable follow-on effects.

None of that is publicly established at this stage.

The current finding is therefore: counterintelligence-relevant individual cases in security-adjacent environments, but no publicly established adversary access pattern.

What remains as the defensible finding?

The cases merit security review, not speculative compression. Several individuals had real technical or institutional relevance. Some cases remain unresolved. The public evidentiary record does not currently support a coordinated abduction, killing or defection campaign against US scientists.

The intelligence relevance lies not in a proven attack on US research. It lies in the transformation of real security proximity into a speculative situation picture. Heterogeneous individual cases are shaped by media compression, political amplification and online speculation into a threat narrative that the known facts do not currently support.

References

House Oversight Committee
Official committee statement on requests for information from NASA, FBI, DOE and the Department of Defense regarding public reports on dead and missing scientists.
oversight.house.gov

CBS News
Reporting on FBI review activity concerning deaths and missing-person cases linked to government laboratories and security-relevant research environments.
cbsnews.com

Scientific American
Analytical context on pattern formation, statistical clustering and the limits of inference from large populations.
scientificamerican.com

MIT News
Institutional obituary and professional profile of Nuno Loureiro, including his role at the Plasma Science and Fusion Center.
news.mit.edu

FBI Boston
Official investigative statement on Claudio Manuel Neves Valente and the absence of a publicly identified terrorism or broader conspiracy nexus.
fbi.gov

Caltech
Institutional obituary and scientific profile of Carl Grillmair and his work at Caltech/IPAC.
caltech.edu

Los Angeles County District Attorney
Official criminal-case information on the murder charge in the Carl Grillmair case.
da.lacounty.gov

The Guardian
Reporting on the Carl Grillmair and Melissa Casias cases, including publicly known case context and the discovery of remains in New Mexico.
theguardian.com

People
Reporting on the Jason Thomas case, including the recovery of remains and public statements on the absence of suspected foul play.
people.com

Snopes
Fact-checking analysis of claims linking multiple dead or missing US scientists to one common pattern.
snopes.com

Glossary

Counterintelligence
Measures used to detect, assess and counter foreign intelligence activity, including recruitment, technology collection, covert access and influence operations.

OSINT
Open-source intelligence. Analysis based on publicly available information, including official statements, court records, media reports, institutional profiles and public databases.

Institutional Proximity
A person’s connection to a sensitive organisation or facility. This does not automatically prove access to classified programmes or security-relevant technical knowledge.

Operational Relevance
The practical value of a person, role or facility for intelligence collection, military assessment, technology acquisition or security planning.

Access Pattern
A recurring method or indicator showing how an actor gains physical, digital or human access to a target environment.

Dual-Use Technology

Technology with both civilian and military applications, including advanced materials, sensors, space propulsion systems and high-performance computing. 

Materials Science
A technical field focused on the structure, properties and use of materials such as metals, alloys, ceramics and composites.

Superalloys
High-performance alloys designed for extreme heat, pressure and mechanical stress. They are relevant for jet engines, rocket components and turbines.

Special Programs
Sensitive or compartmentalised military, intelligence or defence-related programmes. The exact meaning depends on institutional context.

Exfiltration Relevance
The potential value of a person’s knowledge, data access or programme experience for transfer, removal or exploitation by a hostile actor.

UAP
Unidentified Anomalous Phenomena. In this text, UAP narratives matter because they amplify speculation around some cases.

Narrative Compression
The retrospective compression of separate, heterogeneous events into an apparently coherent pattern or threat narrative.

Base-Rate Argument
A statistical caution: in large populations, rare events may appear clustered by chance. This does not disprove a pattern, but it prevents premature conclusions.