None
EN
Comment on OpenAI’s New o3/o4-mini Models Add Invisible Characters to Text, Sparking Watermark Debate by John
['Markus Kasanmascheff', 'Markus Has Been Covering The Tech Industry For More Than Years. He Is Holding A Master S Degree In International Economics', 'Is The Founder', 'Managing Editor Of Winbuzzer.Com.']
Comments for WinBuzzer
Reports surfaced suggesting these newer models embed invisible characters in their text, sparking a debate about whether OpenAI implemented a subtle text watermarking system or if the models are simply exhibiting learned, albeit sometimes problematic, typographical habits.
The method involves a simple find-and-replace to remove the characters, a process Rumi demonstrated in a video.
If accurate, this reframes the finding from a deliberate tracking mechanism to a quirk of the models’ advanced text generation, though the unusual characters could still inadvertently flag text during naive checks.
OpenAI even developed, but paused the rollout of, a linguistic pattern-based text watermarking tool in mid-2024 due to accuracy and bypass concerns.
Beyond Watermarks: Reliability Questions LingerThis specific debate adds to a growing list of observations about the o3 and o4-mini models.
['o4mini'
'o3o4mini'
'watermark'
'watermarking'
'openais'
'rumi'
'characters'
'debate'
'invisible'
'text'
'standard'
'o3'
'ai'
'models'
'sparking'
'openai'
'add']