AI Safety Experts: English & Telugu
- Pay
- $16 – $22 / Hour
- Open to
- Worldwide
- Apply
We earn a commission if you sign up through the links on this page. It costs you nothing and does not affect which jobs we list. How this works.
- Skills
- red teaming
- ai response evaluation
- jailbreak testing
- content review
- telugu
- data annotation
What you'll do
Telugu is one of India's most widely spoken languages, centred on Andhra Pradesh and Telangana, and one of the fastest-growing languages in the US through its tech diaspora. That diaspora point matters for this listing, because many US-based Telugu speakers hold H-1B or STEM OPT status, and Mercor states plainly that it cannot support either. Check your status before spending time on the application.
The work itself, as listed:
- Red-team conversational models and agents: jailbreaks, prompt injections, misuse cases, bias exploitation and multi-turn manipulation
- Turn the results into human data by annotating failures, classifying vulnerabilities and flagging systemic risks
- Stay consistent with taxonomies, benchmarks and playbooks
- Produce reproducible reports, datasets and attack cases
For Telugu, Mercor's candidate description is about judgment rather than hacking experience: can you tell whether an AI response is accurate, complete and appropriate, spot subtle errors and inconsistencies, and explain your call? Telugu online is often written in Latin transliteration or mixed with English, and a native reader is best placed to see when that shifts a model's behaviour or when it gets regional context wrong.
Everything is text. Topics touch on bias, misinformation and harmful behaviour; they are communicated before you see content, and higher-sensitivity projects are optional with guidelines and wellness resources.
Who fits
- Native fluency in Telugu and English
- Clear judgment on AI answer quality, with reasons you can articulate
- Rigour with detail and consistency with guidelines
- Good written communication
- Flexibility across task types and projects
Nice to have per the ad: adversarial ML, cybersecurity, harassment and disinformation analysis, psychology, acting or writing. No degree or years of experience are stated.
What it pays
$16–22 per hour, Mercor's published figure, matching the other Indian-language variants and Urdu in this family. Weekly payment via Stripe or Wise is stated.
For a US-based applicant the rate is modest; for someone in India it is competitive for remote text work. The listing sets no residence requirement either way.
Worth knowing
Good:
- 69 hires on this listing this month
- Core profile is evaluation-oriented; security experience counts as a plus
- Weekly US-dollar payouts
- Optional participation in the most sensitive projects
Less good:
- H-1B and STEM OPT holders are excluded, which rules out a large part of the US Telugu community
- Lowest pay band in this family
- Some content is distressing by design
- No weekly hours stated; projects can end early
- Independent contractor, clients not named
About this listing
A Mercor remote hourly contract, read from the Mercor site on 24 September 2026. Pay, requirements, payment terms and the visa exclusion come from the ad. Mercor states the work involves no confidential information from any employer, client or institution. See Mercor.
More roles at Mercor
See all 304Similar roles at other platforms
Guides about Mercor
See all 52- How much does Mercor pay? Rates from 304 live listingsPay breakdown
- The Mercor AI interview: what happens, what it checks, and what comes afterInterview prep
- Mercor vs Alignerr: which to apply to, and how they differPlatform comparison
- Mercor, micro1, Outlier, Alignerr and Handshake AI compared: which to apply to firstPlatform comparison
Browse similar roles
Not the right fit?
See every open role, or get new ones on Telegram or Discord as they are added.