The committee — Entropy Advisors, MinistroDolar, the Arbitrum Foundation and OpCo, per the proposal — deemed all three cases “high-severity,” defined in the original program proposal as large-scale, deliberate misuse including fabricated deliverables and theft.
As of Sept. 2, the Watchdog Program had received 90 reports, recovered about 532,000 $ARB for the DAO and distributed roughly 268,000 $ARB in reporter bounties, according to the proposal.
Ineligible Users And Self-Farming
Good Entry was allocated 200,000 $ARB in the first round of the Short-Term Incentive Program, which was an up-to-50-million-$ARB push that the DAO approved in late 2023 to incentivize DeFi activity on the chain.
On-chain analysis found 142,839 $ARB of Good Entry’s grant went to 1,032 ineligible users during and after the STIP period, alongside what the committee describes as evidence of self-farming by wallets connected to the team’s own addresses. Asked for clarification, the team “refused to cooperate,” the proposal says.
Good Entry sold protected perpetuals, built on a thesis of pricing gamma in Uniswap LP tokens. It sunset that product on June 28, 2024 after “more than $245m volume,” announced a memecoin launcher on Arbitrum the same day, then stopped development altogether on July 22 in a post-mortem headed “The Dream Ends.” The team blamed liquidity: “Required swaps led to slippage during open/close txs,” it wrote, and “whales prefer principal protected products.” The account has not posted since. DefiLlama still shows about $39,000 in the protocol.
Seventy-Five Thousand To Base
In Limitless’s case, the team swapped its entire 75,000 $ARB LTIPP allocation into USDC and bridged the proceeds to Base, “removing grant funds from the Arbitrum ecosystem entirely,” the committee wrote, classifying it as suspected theft.
The protocol was a liquidation-free, oracle-free margin trading venue, not the Base prediction market of the same name. It launched on Arbitrum on Aug. 6, 2024, and its most recent post is dated Aug. 22 that year. DefiLlama marks it deprecated, with about $2,300 left across Arbitrum and Base. The committee says it could not reach anyone on the team.
From ApolloX To Aster
The largest case is the one where the counterparty still exists. APX Finance was approved for 525,000 $ARB in LTIPP, the follow-on program to STIP, to grow its perpetuals DEX on Arbitrum. The committee found 239,714 $ARB of that spread across three overlapping problems: funds left unspent in team treasury addresses rather than moved to distributor contracts, late distributions to those contracts, and a sybil cluster traced on-chain to team addresses.
Here too, the committee says it could not reach the team. APX’s operating business, meanwhile, is intact. The project announced a merger with Astherus on Dec. 5, 2024, three months after LTIPP incentive streams were required to end, and Astherus rebranded to Aster in March 2025. “Aster (formerly Astherus) is the marriage of two of the strongest products and teams in the $BNB Chain ecosystem,” Aster chief executive Leonard said in an AMA recap published that April. The @APX_Finance account, with 176,000 followers, now tells users to convert APX to ASTER and follow Aster for product updates.
Aster carries $814 million in TVL and did $1.9 billion in perpetuals volume over the past 24 hours, per DefiLlama. Nearly 70% of that value sits on $BNB Chain; about $51 million remains on Arbitrum. Aster launched a privacy-focused Layer 1 in March 2026.
Snapshot Is Final
The DAO approved the Watchdog Program in an on-chain vote that ran May 11-26, 2025 and executed on May 27, with 189.97 million $ARB in favor, 116,140 against and 21.13 million abstaining.
Entropy Advisors had proposed it the previous December with a 400,000 $ARB budget, paying reporters a base bounty tiered by severity — 1,000, 10,000 or 30,000 $ARB — plus 5% of funds recovered, deducted from the recovery and capped by severity at up to $100,000 per report. Every case the committee validated in its April retrospective traced back to STIP, STIP Bridge or LTIPP.
Each project has until Sept. 10 to argue its case as a reply on the forum thread. Six days after publication, the thread carries no replies.
If the explanations do not satisfy the committee and the funds are not returned within the same week, three separate Snapshot votes go up, one per project, asking whether the project “including all founders, current team members, and affiliated contributors” should be permanently banned from all future ArbitrumDAO programs. For projects no longer operational, the ban applies only to founders.
There is no on-chain component. The proposal states that each Snapshot vote “will serve as the final decision,” because implementing a ban requires no on-chain action. The dates are tentative.
The DAO has spent much of 2026 arguing about what it gets for its money, including a $45 million funding request from the Arbitrum Foundation that delegates challenged as spending above DAO revenue. Three ban votes over 457,553 $ARB are a small line in that argument, and the only one that names people.
