The exact text we saved

This is the version exactly as it was saved, character for character. The SHA-256 hash below is a 64-character code worked out from that text. Run the same text through SHA-256 and you get the same code; change a single letter and the code comes out completely different. Each version also stores the previous version’s hash, which chains them together so no version can be swapped out quietly.

TRUVACE RECORD VERSION
record: TRV-2026-2502
version: 1
kind: certified
reason: Certified into the record
timestamp: 2026-11-12T07:43:59.435549Z
status: published
lens: p_space
sector: lifestyle
headline: People offload empathy when they think AI can give it instead
dek: Giving empathy is good, but providing emotional support is hard work. Empathic AI gives cheap and easy support, but may be ethically costly. When AI provides empathy, people may feel less compelled to support others. Four studies provide evidence for “empathy offloading”: when people see AI as capable of empathy, they become less willing to empathize with the suffering of others, even without realizing it. When AI seems more capable of empathy, people—both with correlations (Study 1) and causal experiments (Stud…
gain_title: (none)
problem_title: In four studies, people who saw AI as capable of empathy were less willing to empathize with others' suffering, and learning that someone used an LLM for support led to fewer empathic responses and more social judgment.
trace_subject: (none)
gain_reading: (none)
gain_evidence: (none)
problem_reading: In four studies, people who saw AI as capable of empathy were less willing to empathize with others' suffering, and learning that someone used an LLM for support led to fewer empathic responses and more social judgment.
problem_evidence: when people see AI as capable of empathy, they become less willing to empathize with the suffering of others, even without realizing it | offer fewer empathic responses and more social judgment
quick_read: A paper in Computers in Human Behavior reports four studies on what the authors call "empathy offloading": when people see AI as capable of empathy, they become less willing to empathize with the suffering of others, even without realizing it.

The first two studies tested the link directly. A correlational study (Study 1) and a causal experiment (Study 2) both found that the more capable of empathy AI seemed, the more empathy offloading people reported.

The last two studies looked at how people treat others who have used AI for support. When people learned that someone had received emotional support from large language models, they engaged less with that person's mental states (Study 3) and gave fewer empathic responses and more social judgment (Study 4). The authors warn of empathic devaluing and atrophy, moral deskilling and societal fraying, which are possible implications rather than measured outcomes.
limitation: The authors describe empathic devaluing and atrophy, moral deskilling and societal fraying as possible implications, not outcomes the four studies measured.
tag: Evidence-backed problem
key_points: The paper reports four studies on "empathy offloading", the tendency to hand off empathy to AI. | Study 1 was correlational: people who saw AI as more capable of empathy reported more empathy offloading. | Study 2 was a causal experiment that also found more reported empathy offloading when AI seemed more capable of empathy. | In Study 3, learning that someone had received support from LLMs led people to engage less with that person's mental states. | In Study 4, people offered fewer empathic responses and more social judgment toward someone who had received LLM support. | The authors say the effect can occur without people realizing it. | The authors name possible implications: empathic devaluing and atrophy, moral deskilling and societal fraying.
rundown: The authors frame emotional support as hard work and empathic AI as a cheap, easy alternative that may carry an ethical cost. Their concern is that when AI supplies empathy, people may feel less compelled to support others.

The evidence builds from correlation to causation to behavior. Study 1 found an association between perceiving AI as empathic and offloading empathy, and Study 2 tested the effect experimentally. Studies 3 and 4 moved from self-reported attitudes toward responses to others who had used LLMs for support.

The authors say empathic AI might provide emotional support at any time, but may weaken ethical and emotional commitments to fellow human beings. They list empathic devaluing and atrophy, moral deskilling and societal fraying as possible implications of offloading emotional support.
sources:
- peer_reviewed | Computers in Human Behavior | https://doi.org/10.1016/j.chb.2026.109240 | 2026-10-28
prev: 0000000000000000000000000000000000000000000000000000000000000000

SHA-256 hash

4e540ed10aac53e9c5375a7f126acc4693067bb9ad7558e6b0850455af436ba1

Previous version’s hash

0000000000000000000000000000000000000000000000000000000000000000