RISK & LIABILITY
High Quality Translation and Interpretation Services
What are the risks and liabilities in poor quality translation and interpreting services?
Do you want to avoid serious risks involved in poor quality translation and interpreting services?
- Lawsuits
- Monetary losses
- Malpractice – Any profession
- Inefficient resource utilization
- Delay in product/campaign launch
- Wasted mistranslated marketing campaigns
- Misdiagnosis/delayed diagnosis/medical errors with clinical consequences
- Barriers to access public services (government, legal, education, health, social services)
What are the benefits of high quality translation & interpreting services?
Do you want to have peace of mind with the benefits of Multi-Languages high quality translation & interpreting services?
Safety
- Avoid misunderstandings
- Increased safety & efficiency
- Clear accurate communication
- Strict enforcement of Professional Code of Ethics
- Risk management – communication issues are one of the highest sources of professional liability claims
Connection and engagement
- Improved service performance
- Greater confidence from your clients
- Connect with your clients in their own language
- Higher response and participation in your program
- Professional communication assists with the provision of client centered approach
Business Relationship
- Certified agency
- Save time and money
- Outstanding customer service
- Full Quality Assurance process
- Professional confidentiality guaranteed
Legal Cases
- District sued over translation services for families of students with disabilities
- Famous 1980 Florida Case of Willie Ramirez: Language, Culture and Medical Tragedy
- Welfare Agency Is Sued Over Translation Service
- Spanish interpreter botched 9-1-1 translation, sent ambulance to wrong address, $3 million suit claims
A $3 million wrongful death lawsuit accuses a 9-1-1 Spanish-language interpreter of botching the translation of an address and sending an ambulance to the wrong location as a 25-year-old woman was gasping for air.
- Millions of Americans Are Getting Lost in Translation During Hospital Visits
- YPD routinely fails to translate non-English reports: lawsuit
- 65-Year-Old Woman with an Incorrect Operation on the Left Hand
There was also a problem involving knowledge-based behavior: the ability of the surgeon to speak the patient’s language (and the inability of the other team members to do so) allowed for a nonstandard solution (surgeon acting as interpreter) that effectively shut out other team members from full participation in the informed-consent process. Since the replacement staff members were unable to verify communication between the physician and their patient, a misunderstanding resulted, in which the nurse thought that a conversation between the patient and the surgeon represented a time-out.
Translation Errors AI Risk
SAMPLES
To understand the future of AI, take a look at the failings of Google Translate
March 10, 2025
https://techxplore.com/news/2025-03-future-ai-google.html
The cruellest language barrier: how AI translation is letting down asylum seekers
Feb 18, 2025
Translators’ order cautions public about risks of using artificial intelligence for translation
Oct 16, 2024
Judge Fines Two Lawyers For Using Fake Cases From ChatGPT
June 22, 2023
RESOURCES
Lola’s presentation on AI and Ethics concerning language services
https://docs.google.com/presentation/d/1qeWJkSzHJMr2xGTN2h1O9Fq8NRgdYW2GVLzTOB4IWWQ/edit?usp=sharing
Global AI Ethics and Governance Observatory
https://www.unesco.org/ethics-ai/en?hub=32618
Common AI Translation Errors
- Mistranslation of Context-Specific Phrases
AI often struggles with words that have multiple meanings depending on the context.
- Example:
- Source: “He is charged with battery.”
- AI Translation: “Él está cargado con una batería.” (Spanish — literally: “He is loaded with a battery”)
- Correct: “Está acusado de agresión.”
- Source: “He is charged with battery.”
- Cultural and Idiomatic Misinterpretation
AI fails to recognize idioms or colloquial expressions.
- Example:
- Source: “Kick the bucket”
- AI Translation: “Patear el balde” (Literal in Spanish)
- Correct: “Morir” (To die)
- Source: “Kick the bucket”
- Gender and Pronoun Confusion
Some languages have gendered grammar that AI misinterprets, especially when translating from gender-neutral languages like English.
- Example:
- English: “The doctor said they will call.”
- AI in French: “Le médecin a dit qu’ils appelleront.” (Defaults to plural masculine, loses singular reference)
- English: “The doctor said they will call.”
- Formality and Register Errors
AI may default to the wrong level of politeness, which can be offensive in languages like Korean, Japanese, or German.
- Example (Japanese):
- Casual form used in a professional email
- AI Translation: “じゃあね” (“See ya”) instead of “失礼いたします” (“Excuse me formally”)
- Casual form used in a professional email
- Inconsistency in Terminology
AI doesn’t always maintain consistency in specialized terms across a document—critical in legal, medical, and technical translations.
- Example:
- “Plaintiff” translated as both “demandante” and “parte actora” in the same Spanish legal document
- “Plaintiff” translated as both “demandante” and “parte actora” in the same Spanish legal document
- Hallucinations / Fabricated Content
AI occasionally invents content that wasn’t in the source text.
- Example:
- Source: “The report outlines three key findings.”
- AI Translation: Adds extra “findings” not present in the original
- Source: “The report outlines three key findings.”
- False Fluency (Looks Correct, But Isn’t)
The translation sounds fluent but conveys the wrong meaning subtly.
- Example:
- Original: “He filed for bankruptcy.”
- AI Translation: “Presentó documentos para la bancarrota.”
- Correct: “Se declaró en bancarrota.”
- Original: “He filed for bankruptcy.”
- Ignoring Industry-Specific Jargon
AI may mistranslate technical or domain-specific language unless finely tuned.
- Example (Medical):
- “Stroke” mistranslated as “golpe” (a hit) instead of “accidente cerebrovascular”
- “Stroke” mistranslated as “golpe” (a hit) instead of “accidente cerebrovascular”
- Language Biases & Stereotypes
AI trained on biased data may reflect stereotypes.
- Example:
- Translates “The nurse” in masculine contexts and “The engineer” in feminine contexts, or vice versa depending on language pair and data
- Translates “The nurse” in masculine contexts and “The engineer” in feminine contexts, or vice versa depending on language pair and data
- Failure to Adapt to Local Varieties
AI may default to one dialect and ignore localization needs.
- Example:
- Spanish: Using “ordenador” (Spain) for “computer” when “computadora” (Latin America) is appropriate
- Spanish: Using “ordenador” (Spain) for “computer” when “computadora” (Latin America) is appropriate