Pump.fun has rolled out an automated buyback-and-burn system, yet the PUMP token kept sliding. Market data in the source shows PUMP fell 8.63% to $0.00196, while Bitcoin dropped 2.53% to $70,481 and the total crypto market cap slipped to $2.4 trillion. The feature launch was substantial, but short-term price action remained tied to a broader market sell-off.
AI agent revenue is now linked directly to token buybacks
The new setup connects onchain agents to the token economy in a more explicit way. Developers can deploy AI agents tied to products or trading activity, and when those agents generate revenue, a preset share of profits is used to buy back and burn the related token. That lowers supply and turns agent revenue into direct buy-side demand. According to the source, builders can use tools such as Claude Code and choose their own buyback percentages.
The idea is to close a gap seen in earlier AI agent projects, where agent activity could grow without creating a clear benefit for token holders. Here, revenue, buybacks, and token burns sit in one system.
No project-specific negative trigger was cited
The report says the token is not reacting to a fresh exploit, protocol failure, or other isolated bad news. In fact, the protocol reportedly bought back $1.27 million worth of tokens over the last 24 hours. That points more to profit-taking and risk-off sentiment across the market than to a breakdown in the new feature set.
The source describes the move as a kind of negative beta reaction, meaning the token is behaving like a more volatile version of Bitcoin. Once Bitcoin moved lower, PUMP weakened even more. The upgrade did not disappear; it was simply overshadowed by macro price pressure inside crypto.
$0.0019 support is the near-term level in focus
Near-term levels in the source are straightforward. If Bitcoin can hold above $70,000, PUMP may find support near $0.0019 and then retest $0.0021 as resistance. If the wider market keeps falling, the next area to watch is $0.0018.
Daily buyback volume is another metric to monitor. If AI agents begin generating more revenue, automated buybacks could add stronger buying pressure over time. For now, the mechanism looks more like a long-term support layer than a tool capable of reversing a market-wide slide on its own.