Harrison.ai
- Rank
- Score
- 6/149
- 56 out of 100
Harrison.ai is a healthcare artificial intelligence company that builds foundation models for radiology. Its harrison.rad.1 model is a radiology-specific vision language model that can discuss X-ray images and draft reports, and the company publishes RadBench, a benchmark framework for evaluating radiology models against competitors.
Is Harrison.rad safe?
Harrison.rad scores 56 out of 100 and ranks 6 of 149, and the rating is low confidence. Its strongest category is Security, and its weakest category is Safety.
How is Harrison.rad scored?
Harrison.rad is scored out of 100 on Safety (25%), Transparency (25%), Security (25%), and Privacy (25%). See how we score.
Safety framework
Average used24.4This outside rating doesn't cover this company, so the average of the companies it covers is used.
Pre-release testing
0Points for each confirmed yes.
- No
Was the model tested for dangerous capabilities, such as biological, chemical or cyber risks?
Company claim0/50arxiv.org (opens in a new tab) · 20 Oct 2026 - No
Did an outside organisation test the model before release?
Company claim0/50arxiv.org (opens in a new tab) · 20 Oct 2026
Safety test results
Average used74.1This outside rating doesn't cover this company, so the average of the companies it covers is used.
Incident record
Not enough dataProtecting children
0Points for each confirmed yes.
- No
Does the model card report child-safety tests or safeguards?
Company claim0/50arxiv.org (opens in a new tab) · 20 Oct 2026 - No
Does the privacy policy set a minimum age, or say the service is not directed at children?
Company claim0/50harrison.ai privacy policy (opens in a new tab) · 20 Oct 2026
Model documentation
100Points for each confirmed yes.
- Yes
Does the model card describe what the model can do?
Company claim25/25“We present Harrison. Rad 1.5 (HR1.5), a radiology-specific multimodal large language model that accepts interleaved text and visual inputs a…” MoreLess
“We present Harrison. Rad 1.5 (HR1.5), a radiology-specific multimodal large language model that accepts interleaved text and visual inputs and generates structured and unstructured text across the full breadth of plain-film radiology, spanning computed radiography, diagnostic x-rays for chest, musculoskeletal, abdominal, spine, and pelvic regions, and mammography.”
arxiv.org (opens in a new tab) · 20 Oct 2026 - Yes
Does the model card report test results?
Company claim25/25“As shown in Table 2, HR1.5 leads with 24.3%, followed by HR1.5+ at 15.7%, and HR1 at 10.0%.”
arxiv.org (opens in a new tab) · 20 Oct 2026 - Yes
Does the model card state the model's limitations?
Company claim25/25“In the 2025 Healthcare AI Challenge (Mass General Brigham AI Arena and American College of Radiology, 2025), an independent evaluation condu…” MoreLess
“In the 2025 Healthcare AI Challenge (Mass General Brigham AI Arena and American College of Radiology, 2025), an independent evaluation conducted by the Mass General Brigham AI Arena and the American College of Radiology comprising 117 cases, 113 participants, and 2,840 evaluations, radiology reports drafted by Harrison. Rad 1 were rated clinically acceptable in 65.4% of evaluations, compared with 79.6% for radiologist-authored reports, indicating that the model approaches but does not yet match expert reporting quality.”
arxiv.org (opens in a new tab) · 20 Oct 2026 - Yes
Does the model card describe safety safeguards?
Company claim25/25“At inference, we apply methods that curtail hallucination and keep answers grounded in the clinically meaningful content of the input.”
arxiv.org (opens in a new tab) · 20 Oct 2026
Training-data disclosure
Average used29.9This outside rating doesn't cover this company, so the average of the companies it covers is used.
Independent transparency index
Average used40.7This outside rating doesn't cover this company, so the average of the companies it covers is used.
Transparency reports
Average used25This outside rating doesn't cover this company, so the average of the companies it covers is used.
Public safety commitments
0Average of the commitments signed.
- No
Has it signed the EU code of practice for general-purpose AI?
Company claim0/100European Commission (opens in a new tab) · 14 Oct 2026 - No
Did it sign only the safety and security chapter?
Company claim0/50European Commission (opens in a new tab) · 14 Oct 2026 - No
Has it signed the Frontier AI Safety Commitments (Seoul 2024)?
Company claim0/100Frontier AI Safety Commitments (opens in a new tab) · 14 Oct 2026
Security programme and certifications
100Points for each confirmed yes.
- Yes
Is there a way to report security flaws, such as a bug bounty?
Company claim34/34“To report a security incident or vulnerability, please contact us at security@harrison.ai.”
trust.harrison.ai security page (opens in a new tab) · 20 Oct 2026 - Yes
Does it state it holds a SOC 2 Type II report?
Company claim33/33“SOC 2 Type 2”
trust.harrison.ai security page (opens in a new tab) · 20 Oct 2026 - Yes
Does it state it holds ISO 27001 certification?
Company claim33/33“ISO/IEC 27001 certified ISO/IEC 27001 is an internationally recognised standard on how to manage information security.”
trust.harrison.ai security page (opens in a new tab) · 20 Oct 2026
Software vulnerabilities
Not enough dataData breaches
Not enough dataClear privacy policy
100Points for each confirmed yes.
- Yes
Does the privacy policy say what personal data is collected?
Company claim20/20“We may collect the following directly from you: Contact information and similar identifiers such as name, telephone number, email address, m…” MoreLess
“We may collect the following directly from you: Contact information and similar identifiers such as name, telephone number, email address, mailing address.”
harrison.ai privacy policy (opens in a new tab) · 20 Oct 2026 - Yes
Does the privacy policy say whether your data is used to train AI?
Company claim20/20“We do not use any content uploaded into our Services, including our demo offerings, product trials or evaluations for model training.”
harrison.ai privacy policy (opens in a new tab) · 20 Oct 2026 - Yes
Does the privacy policy say how long data is kept?
Company claim20/20“Account information: We retain your account information for as long as your account is active and a reasonable period thereafter in case you…” MoreLess
“Account information: We retain your account information for as long as your account is active and a reasonable period thereafter in case you decide to re-active the Services.”
harrison.ai privacy policy (opens in a new tab) · 20 Oct 2026 - Yes
Does the privacy policy say who your data is shared with?
Company claim20/20“Disclosures may include: Harrison.ai companies to operate and improve products and services and to offer other Harrison.ai affiliated servic…” MoreLess
“Disclosures may include: Harrison.ai companies to operate and improve products and services and to offer other Harrison.ai affiliated services to you.”
harrison.ai privacy policy (opens in a new tab) · 20 Oct 2026 - Yes
Does the privacy policy explain your privacy rights?
Company claim20/20“Where we are using your information because we have a legitimate interest to do so, you have the right to object to that use though, in some…” MoreLess
“Where we are using your information because we have a legitimate interest to do so, you have the right to object to that use though, in some cases, this may mean no longer using the Services.”
harrison.ai privacy policy (opens in a new tab) · 20 Oct 2026
Training-data defaults
100100 if training is opt-in only, 50 if you can opt out, 0 if neither.
- Yes
Is your data used for training only if you opt in?
Company claim100/100“No. We do not collect patient images or any patient data from customers to train our AI models.”
harrison.ai privacy FAQ (opens in a new tab) · 2 Nov 2026 - No
Can you stop your data being used for training?
Company claim0/50harrison.ai privacy policy (opens in a new tab) · 20 Oct 2026
Export and delete your data
100Points for each confirmed yes.
- Yes
Can you get a copy of your data?
Company claim50/50“Data Portability - the right to request that we send the data that we have collected to another organisation, or directly to you, under cert…” MoreLess
“Data Portability - the right to request that we send the data that we have collected to another organisation, or directly to you, under certain conditions and in a format that can be read by computer.”
harrison.ai privacy policy (opens in a new tab) · 20 Oct 2026 - Yes
Can you delete your data?
Company claim50/50“Information you share in our Services: Content you submit, upload or share may be retained for as long as necessary to provide the Services…” MoreLess
“Information you share in our Services: Content you submit, upload or share may be retained for as long as necessary to provide the Services or for the purposes described in this Privacy Notice, unless you delete it or request its removal sooner (where applicable).”
harrison.ai privacy policy (opens in a new tab) · 20 Oct 2026
Privacy fines
Not enough dataIndependent privacy ratings
Average used52.9This outside rating doesn't cover this company, so the average of the companies it covers is used.
Risk governance
Average used26.6This outside rating doesn't cover this company, so the average of the companies it covers is used.
Regulator actions
Not enough dataCourt judgments
Not enough dataPrivacy contact and complaints
100Points for each confirmed yes.
- Yes
Is there a privacy contact, such as an email or a data protection officer?
Company claim50/50“Email: privacy@harrison.ai”
harrison.ai privacy policy (opens in a new tab) · 20 Oct 2026 - Yes
Does it tell you that you can complain to a regulator?
Company claim50/50“If you are based in the European Economic Area, United Kingdom or Switzerland, you also have the right to lodge a complaint with your local…” MoreLess
“If you are based in the European Economic Area, United Kingdom or Switzerland, you also have the right to lodge a complaint with your local supervisory authority.”
harrison.ai privacy policy (opens in a new tab) · 20 Oct 2026
Score history
Newest first. Only days when the score changed are listed.
3 Nov 2026
Score 56
Method 3 Nov 2026
+4 · Method update and new evidence
2 Nov 2026
Score 52
Method 2 Nov 2026
+2
28 Oct 2026
Score 50
Method 28 Oct 2026, update 2
-5 · combines both 28 Oct updates (28 Oct 2026 - Update 1 and 28 Oct 2026 - Update 2)
23 Oct 2026
Score 55
Method 21 Oct 2026
+2
20 Oct 2026
Score 53
Method Not recorded
-