Draft, not yet released. In stewardship review. Not indexed until the first approved release.

Risks and capability

Four robotics risk objects open the portfolio. Order carries no rank. Each is read across driver, exposure, vulnerability, capacity, cascade and consequence.

6 min read0 figures1 table23 notesNR-XTECH-ROBOTICS-GLOBAL-2026-BASELINE

Contents
  1. ·Edition home
  2. ·Executive summary
  3. 1Robots are infrastructure and a test of institutions
  4. 2Key findings
  5. 3Systems and sectors at baseline
  6. 4Risks and capability
    1. 4.1 Risk objects
    2. 4.2 Capability ladder
  7. 5Humanoid pilots meet a missing safety standard
  8. 6Readings from the other desks
  9. 7The 2027 agenda
  10. 8Systems in depth
  11. AMethodology
  12. BData appendix
  13. CReferences and record
  14. ·Complete edition, for print

4.1Risk objects

Each risk is read across the chain from hazard to consequence, with its status and evidence-quality level.

  1. InferredEQL 3RSK-XT03-GOV-001

    Rules lag deployment

    Driver
    humanoid and mobile platforms entering workplaces ahead of finished standards1532
    Exposure
    workers in factories and warehouses; the public near driverless vehicles
    Vulnerability
    draft standards; rules that differ between jurisdictions
    Capacity
    machinery rules, vehicle regulation and robot safety standards172231
    Cascade
    GOV → INF, FIN
    Horizon
    1 to 5 years

  2. InferredEQL 3RSK-XT03-MIN-002

    Magnet and actuator supply chokepoint

    Driver
    export licensing on rare earth elements and magnets2141
    Exposure
    actuator makers; humanoid and legged platforms
    Vulnerability
    few suppliers; concentrated processing3940
    Capacity
    alternative motor designs; stocks; recycling
    Cascade
    MIN → INF
    Horizon
    1 to 5 years

  3. Expert inputEQL 2RSK-XT03-INF-003

    Thin field safety evidence

    Driver
    one study of road vehicles and no consolidated data elsewhere530
    Exposure
    planners who treat company statistics as settled
    Vulnerability
    incident reporting that differs by sector
    Capacity
    federal crash reporting and independent studies530
    Cascade
    INF → DAT → GOV
    Horizon
    1 to 5 years

  4. InferredEQL 2RSK-XT03-FIN-004

    Local labour adjustment

    Driver
    rising fleets in concentrated markets120
    Exposure
    workers in routine manufacturing and logistics roles
    Vulnerability
    regions with one dominant employer or sector
    Capacity
    skills programmes; social protection; none bound here
    Cascade
    FIN → HLT, DAT
    Horizon
    5 to 10 years

4.2Capability ladder

Existence is not readiness. Each of the fifteen capability functions is placed on the ladder by the strongest public evidence for robotics at the global level.

Table 4.1Capabilities on the ladder of evidence: listed, exists, demonstrated, tested
FunctionRungLevelStrongest public evidence at baseline
Prevention and mitigationEXISTSrobot safety standards and machinery rules published3135
Monitoring and observationLISTEDno common tracker of robot incidents outside road vehicles30
Early warningLISTEDnot consolidated across sectors
Preparedness and planningEXISTSEU machinery timetable and UN vehicle regulation1722
Redundancy and reservesLISTEDmagnet and actuator stocks are not documented41
ResponseLISTEDno tested response to a fleet-wide fault
Continuity of critical servicesLISTEDdependence of logistics on fleets not reported
RecoveryLISTEDnot applicable at baseline; no event to recover from
AdaptationLISTEDrules for learning systems are still being written1636
Social protectionLISTEDnot bound here
Finance and insuranceLISTEDan insurance-institute study exists for road vehicles5
Institutional coordinationEXISTSstandards bodies and the UN vehicle forum223132
Technical infrastructureEXISTSlarge installed fleets and driverless services14
WorkforceLISTEDno recent figure bound
Research capacityADEQUATEstrong preprint base on learned control789101112
A capability listed in policy is not a tested one; each rung is the strongest public evidence at baseline.

Notes

Numbers are the edition’s reference numbers, so a note reads the same in every chapter. The full list, with every source and its licence, is in Annex C. Annex C, references

  1. 15Engadget (ed.), Boston Dynamics unveils production-ready version of Atlas robot at CES 2026 (report of 8 January 2026; first customers Hyundai and Google DeepMind; Hyundai plans car-plant use from 2028). Trade press reporting a company plan, weak evidence. Observation year 2026. Accessed 2026-11-03.back
  2. 32UNVERIFIED, not opened: International Organization for Standardization (ed.), ISO 25785-1, safety requirements for industrial mobile robots with actively controlled stability, Part 1: Robots, committee draft in preparation, taken from a search summary of the catalogue entry. ISO, Geneva. Observation year 2025. Accessed 2026-11-03.back
  3. 17European Agency for Safety and Health at Work (ed.), Regulation (EU) 2023/1230 on machinery, legislation page (applies from 20 January 2027; replaces Directive 2006/42/EC; aims to cover autonomous mobile machinery and artificial-intelligence safety modules). EU-OSHA, Bilbao. Observation year 2025. Accessed 2026-11-03.back
  4. 22Connected Automated Driving (ed.), UNECE adopts first-ever global rules allowing fully autonomous vehicles (page of 8 July 2026; adopted 24 June 2026; safety-case approach; about 90 existing UN regulations amended). Trade summary of a UN decision, weak evidence. Observation year 2026. Accessed 2026-11-03.back
  5. 31UNVERIFIED, not opened: International Organization for Standardization (ed.), ISO 10218-1:2025 Robots and robotic devices, safety requirements for industrial robots, Part 1 (third edition, published 5 February 2025), taken from a search summary of the catalogue entry. ISO, Geneva. Observation year 2025. Accessed 2026-11-03.back
  6. 21Adamas Intelligence (ed.), Musk on rare earth export restrictions: Optimus was affected (article of 24 April 2025; Tesla first-quarter 2025 earnings call; arm actuators use permanent magnets). Trade analysis reporting a company statement, weak evidence. Adamas Intelligence, Toronto. Observation year 2025. Accessed 2026-11-03.back
  7. 41UNVERIFIED, not opened: Ministry of Commerce of the People’s Republic of China (ed.), announcement of export controls on seven rare earth elements and related magnets, April 2025, taken from a search summary. MOFCOM, Beijing. Observation year 2025. Accessed 2026-11-03.back
  8. 39UNVERIFIED, not opened: IEA (ed.), Global Critical Minerals Outlook 2026, the page returned an access error; no figure used. International Energy Agency, Paris. Observation year 2026. Accessed 2026-11-03.back
  9. 40UNVERIFIED, not opened: US Geological Survey (ed.), Mineral Commodity Summaries 2026, rare earths, the file returned an access error; no figure used. USGS, Reston. Observation year 2026. Accessed 2026-11-03.back
  10. 5Insurance Institute for Highway Safety (ed.), Waymo’s driverless cars crash less often than people (news release, 23 July 2026; about 50 million driverless miles, 2021 to 2024; 68% lower crash involvement; 81% fewer injury crashes per mile; four cities). IIHS, Arlington. Observation year 2024. Accessed 2026-11-03.back
  11. 30UNVERIFIED, not opened: National Highway Traffic Safety Administration (ed.), Standing General Order 2021-01 on crash reporting, Third Amended order of 24 April 2025, taken from a search summary; the agency page returned an access error. NHTSA, Washington. Observation year 2025. Accessed 2026-11-03.back
  12. 1International Federation of Robotics (ed.), Five Million Robots now Operate in Factories Globally (World Robotics 2026 press release, Frankfurt, 24 September 2026; installations of more than 600,000 in 2025, up 11%; China 354,000; stock 5 million; forecast 655,000 in 2026 and 806,000 in 2029). International Federation of Robotics, Frankfurt. Observation year 2025. Accessed 2026-11-03.back
  13. 20Acemoglu, D. and Restrepo, P., Robots and Jobs: Evidence from US Labor Markets, NBER Working Paper 23285, March 2017 (US adoption 1990 to 2007). National Bureau of Economic Research, Cambridge. Observation year 2017. Accessed 2026-11-03.back
  14. 35UNVERIFIED, not opened: Regulation (EU) 2023/1230 of the European Parliament and of the Council of 14 June 2023 on machinery, Official Journal of the European Union; the official page returned no content, summary read through the EU-OSHA page. Observation year 2023. Accessed 2026-11-03.back
  15. 16European Commission (ed.), AI Act, regulatory framework page (entered into force 1 August 2024; high-risk dates revised by an omnibus in force 27 July 2026; AI-based safety components of products such as surgical robots treated as high-risk). Directorate-General for Communications Networks, Content and Technology, Brussels. Observation year 2026. Accessed 2026-11-03.back
  16. 36UNVERIFIED, not opened: Regulation (EU) 2024/1689 of the European Parliament and of the Council laying down harmonised rules on artificial intelligence, Official Journal of the European Union, 2024, cited from memory; summary read through the Commission page. Observation year 2024. Accessed 2026-11-03.back
  17. 4Waymo (ed.), Safety Impact hub (page read November 2026; 271.3 million rider-only miles through June 2026; about 95% fewer serious-injury-or-worse crashes than a human benchmark). Company-reported analysis, weak evidence. Waymo, Mountain View. Observation year 2026. Accessed 2026-11-03.back
  18. 7Brohan, A. et al., RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Control, arXiv:2307.15818, submitted 28 July 2023 (company authors; 6,000 evaluation trials). arXiv, Ithaca. Observation year 2023. Accessed 2026-11-03.back
  19. 8Open X-Embodiment Collaboration, Open X-Embodiment: Robotic Learning Datasets and RT-X Models, arXiv:2310.08864, submitted 13 October 2023, revised 14 May 2025 (22 robots, 21 institutions). arXiv, Ithaca. Observation year 2023. Accessed 2026-11-03.back
  20. 9Black, K. et al., pi0: A Vision-Language-Action Flow Model for General Robot Control, arXiv:2410.24164, submitted 31 October 2024, revised 8 January 2026 (company preprint). arXiv, Ithaca. Observation year 2024. Accessed 2026-11-03.back
  21. 10Gemini Robotics Team, Gemini Robotics: Bringing AI into the Physical World, arXiv:2503.20020, submitted 25 March 2025 (company preprint; learning of short-horizon tasks from as few as 100 demonstrations). arXiv, Ithaca. Observation year 2025. Accessed 2026-11-03.back
  22. 11Octo Model Team, Octo: An Open-Source Generalist Robot Policy, arXiv:2405.12213, submitted 20 May 2024 (800,000 trajectories from Open X-Embodiment; nine robot platforms). arXiv, Ithaca. Observation year 2024. Accessed 2026-11-03.back
  23. 12Li, C. et al., BEHAVIOR-1K: A Human-Centered, Embodied AI Benchmark with 1,000 Everyday Activities and Realistic Simulation, arXiv:2403.09227, submitted 14 March 2024 (abstract page; 1,000 activities, 50 scenes). arXiv, Ithaca. Observation year 2024. Accessed 2026-11-03.back

Cite this chapter

Valadbaygi, S. (Ed.). (2026). Chapter 4: Risks and capability. In Exponential Technology Report 2026: Robotics and Autonomous Systems (Baseline Edition 2026, Report No. NR-XTECH-ROBOTICS-GLOBAL-2026-BASELINE) [Draft under review]. The Global Centre for Risk and Innovation (GCRI).

Every citation form of the whole report, BibTeX and RIS included, is on the edition home. Cite this report

ReportsContribute

Send evidence or a correction on Robotics and Autonomous Systems.

Reports is the Consortium’s quarterly publication, written by six desks. This page sits with the Global Centre for Risk and Innovation (GCRI) Systems Evidence desk. What you know about this subject can become evidence, a case or a correction in the next edition, credited to you and published under CC BY 4.0 where it is your own work. Sent to the wrong desk, it is forwarded, not refused.

GCRI Systems Evidence deskExponential Technology

Never send personal data, sensitive infrastructure detail or anything you have no right to share. A contribution is reviewed before use; sending it confers no authorship or editorial rights, and sponsorship never buys a finding.

CC BY 4.0 Free to read. Text and figures: CC BY 4.0. Data: CC0 1.0. Authors keep copyright. Published by GCRI Canada.

The Global Centre for Risk and Innovation

Science · evidence · measurement

Not for profit

Our articles, maps and twins are free to read and free to republish. Readers and partners keep it that way.

Partner with us

Who funds us

GCRI is a not-for-profit institution. To ask who funds a piece of work, use the funding questions form.

Independence

Editors and the expert reviewers named on each article decide what we publish. Funders and partners do not review or approve articles before publication.

Corrections policy

Licence

Articles: CC BY 4.0. Figures, maps, twins and data: CC BY 4.0. Other uses: ask through the permissions form.

Privacy choices