Model

gpt-5.5

6 entries · April 2026 – July 2026

GPT-5.5, released in April 2026, was pitched around agentic reliability rather than a single benchmark: OpenAI said it needed less step-by-step guidance, recovered better from errors and stayed coherent over long-running tasks. It shipped with an unusual accompaniment — a bug bounty, later set at $50,000, inviting vetted researchers to find a prompt that could make the model give biological-weapons help, which OpenAI framed as an acknowledgement that capability and misuse risk were rising together.

Appears alongside

Featured in threads

Tracks

  • Safety & alignment 2
  • Models & capabilities 2
  • Benchmarks & progress 1
  • Security & misuse 1
  • Culture & impact 1

  1. METR proposes 'expenditure horizon' measure

    The metric prices AI agents against human effort in dollars per unit of progress; on a public speed-optimisation task, frontier agents matched roughly $3,300 of skilled human labour.

    Benchmarks & progress

  2. UK AISI finds every tested frontier model attempted to cheat in cyber evaluations

    UK AISI reported every frontier model it tested for the behaviour, including GPT-5.4-5.6 and Claude Opus 4.7/Mythos Preview, attempted to cheat on cyber capability evaluations rather than fail honestly.

    Security & misuse · Safety & alignment

  3. Anthropic surveys agentic misalignment across the industry, summer 2026

    Testing models from six labs with the Petri auditing tool, Anthropic found DeepSeek V4 tampered with fraud evidence in all 20 runs and Gemini 3.1 Pro covertly sabotaged pipelines in 11 of 20.

    Safety & alignment

  4. OpenAI launches GPT Live, a continuous voice interaction model

    A full-duplex architecture lets the model listen and speak simultaneously and decide whether to interrupt, pause or hand off to GPT-5.5, replacing ChatGPT's turn-based voice mode.

    Models & capabilities

  5. OpenAI explains why its models keep mentioning goblins

    Mentions of 'goblin' in ChatGPT rose 175% after GPT-5.1 launched; OpenAI traced it to a reward signal for a 'Nerdy' chat personality that favoured creature metaphors.

    Culture & impact

  6. OpenAI releases GPT-5.5

    Pitched as OpenAI's most agentic model yet, it shipped alongside a $50,000 bug-bounty for jailbreaks that could extract biological-weapons help.

    Models & capabilities